From patchwork Mon Jan 25 19:08:08 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044035 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 84CFDC433E0 for ; Mon, 25 Jan 2021 19:09:16 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 3FAE0216FD for ; Mon, 25 Jan 2021 19:09:16 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 3FAE0216FD Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74298.133525 (Exim 4.92) (envelope-from ) id 1l47EZ-0008VF-4K; Mon, 25 Jan 2021 19:09:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74298.133525; Mon, 25 Jan 2021 19:09:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47EZ-0008V2-02; Mon, 25 Jan 2021 19:09:07 +0000 Received: by outflank-mailman (input) for mailman id 74298; Mon, 25 Jan 2021 19:09:05 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47EX-0008N7-Ft for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:05 +0000 Received: from mail-wm1-x332.google.com (unknown [2a00:1450:4864:20::332]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 1bf78242-ef3a-4bc0-a562-3df55556a015; Mon, 25 Jan 2021 19:08:54 +0000 (UTC) Received: by mail-wm1-x332.google.com with SMTP id i9so560585wmq.1 for ; Mon, 25 Jan 2021 11:08:54 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.52 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:53 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 1bf78242-ef3a-4bc0-a562-3df55556a015 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=T4YlJPXxZbXKgBsejUH07Ya0YvCCETQ+Uu7/Rje4XkQ=; b=M04qiFqNVD66m2MWoGNhNKculhhp/yHquCxkhF/Ppn3D0WfoGDzhIDBd8h11hn8GBM E1J1vt+51knG0U7SapDz5nkpN/Y9PRDxTJbjU4sOBrSikm0I3sLdpAzod+XIzqNJB6ra fYGph/n1+LNnfuBTbq0izXOh21lxJaOgNTdQoL0zv6XR8IkM3YQenw+dJB38iq4Ukv/J 5kN4DS0qH91ANgfxaKV0fM1mDXdHHKBKlfwNqe0G0ZIJdi/xEOELcrZmMUxa88qWPeYZ RlEpXdZrOGu6RuR8r8kwoY4KWJAElpDQJZ7TYizJUkctoRrHQHfla+17BaVggB5CIIjy X/1Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=T4YlJPXxZbXKgBsejUH07Ya0YvCCETQ+Uu7/Rje4XkQ=; b=CPhJNEldIDpVZekZafsz6Flz7PLn+daJg790J1uwizRZo3u9U+E4tmrKt7yP8bBNZt 1ffHdHh6jtH8koUdL+VyVMB8n+Z23xPRa4PnZt9P4aYqw6L+Un6dvtAIWjI7cYpKD4RU Kwki8y7ZPFbcIK9eEABBvVon7o3fbFxB1zbtRsHLNWfa40aq8ELddgztoRywjjoTGxr/ MtuNurtOu0/mMck57Eoa5NNdo3Z6gKK1xSXVPstvB3a8ieIGyvSCfsH9+L1lfK6mA0hC FaCEZbMLGZrN2d8dvNiEasOPtnbEXo9pFbg0kXm8g6dQ8tCfO53mToA37RQoHTUhMBQe Sdng== X-Gm-Message-State: AOAM531gAyVph67g7woRw8omsco4H5SfptMQoU4nMvF0R5uH02HJsk0J yiuggjxOZervLlrEc54QjTZrXxemFYqqYQ== X-Google-Smtp-Source: ABdhPJwhgdaM3c4+MfVVPDpQhYLDJsXNbIjvhJ9mYh3ZxL0/Y4caRqY8tCjRpW10DXvzSzh1GODqAw== X-Received: by 2002:a1c:9844:: with SMTP id a65mr1435985wme.113.1611601733422; Mon, 25 Jan 2021 11:08:53 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 01/22] x86/ioreq: Prepare IOREQ feature for making it common Date: Mon, 25 Jan 2021 21:08:08 +0200 Message-Id: <1611601709-28361-2-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko As a lot of x86 code can be re-used on Arm later on, this patch makes some preparation to x86/hvm/ioreq.c before moving to the common code. This way we will get a verbatim copy for a code movement in subsequent patch. This patch mostly introduces specific hooks to abstract arch specific materials taking into the account the requirment to leave the "legacy" mechanism of mapping magic pages for the IOREQ servers x86 specific and not expose it to the common code. These hooks are named according to the more consistent new naming scheme right away (including dropping the "hvm" prefixes and infixes): - IOREQ server functions should start with "ioreq_server_" - IOREQ functions should start with "ioreq_" other functions will be renamed in subsequent patches. Introduce common ioreq.h right away and put arch hook declarations there. Also re-order #include-s alphabetically. This support is going to be used on Arm to be able run device emulator outside of Xen hypervisor. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Alex Bennée Reviewed-by: Julien Grall Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen Acked-by: Jan Beulich --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch, was split from: "[RFC PATCH V1 01/12] hvm/ioreq: Make x86's IOREQ feature common" - fold the check of p->type into hvm_get_ioreq_server_range_type() and make it return success/failure - remove relocate_portio_handler() call from arch_hvm_ioreq_destroy() in arch/x86/hvm/ioreq.c - introduce arch_hvm_destroy_ioreq_server()/arch_handle_hvm_io_completion() Changes V1 -> V2: - update patch description - make arch functions inline and put them into arch header to achieve a truly rename by the subsequent patch - return void in arch_hvm_destroy_ioreq_server() - return bool in arch_hvm_ioreq_destroy() - bring relocate_portio_handler() back to arch_hvm_ioreq_destroy() - rename IOREQ_IO* to IOREQ_STATUS* - remove *handle* from arch_handle_hvm_io_completion() - re-order #include-s alphabetically - rename hvm_get_ioreq_server_range_type() to hvm_ioreq_server_get_type_addr() and add "const" to several arguments Changes V2 -> V3: - update patch description - name new arch hooks according to the new naming scheme - don't make arch hooks inline, move them ioreq.c - make get_ioreq_server() local again - rework the whole patch taking into the account that "legacy" interface should remain x86 specific (additional arch hooks, etc) - update the code to be able to use hvm_map_mem_type_to_ioreq_server() in the common code (an extra arch hook, etc) - don’t include from arch header - add "arch" prefix to hvm_ioreq_server_get_type_addr() - move IOREQ_STATUS_* #define-s introduction to the separate patch - move HANDLE_BUFIOREQ to the arch header - just return relocate_portio_handler() from arch_ioreq_server_destroy_all() - misc adjustments proposed by Jan (adding const, unsigned int instead of uint32_t) Changes V3 -> V4: - add Alex's R-b - update patch description - make arch_ioreq_server_get_type_addr return bool - drop #include - use two arch hooks in hvm_map_mem_type_to_ioreq_server() to avoid calling p2m_change_entry_type_global() with lock held Changes V4 -> V5: - add Julien's and Paul's R-b - update patch description - remove single use variable in arch_ioreq_server_map_mem_type_completed() - put multiple function parameters on a single line in the header where possible - introduce common ioreq.h right away and put arch hooks declarations there instead of doing that in patch #4 --- --- xen/arch/x86/hvm/ioreq.c | 175 +++++++++++++++++++++++++++++++---------------- xen/include/xen/ioreq.h | 54 +++++++++++++++ 2 files changed, 169 insertions(+), 60 deletions(-) create mode 100644 xen/include/xen/ioreq.h diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 1cc27df..3c3c173 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -16,16 +16,16 @@ * this program; If not, see . */ -#include +#include +#include #include +#include +#include #include -#include +#include #include -#include #include -#include -#include -#include +#include #include #include @@ -170,6 +170,29 @@ static bool hvm_wait_for_io(struct hvm_ioreq_vcpu *sv, ioreq_t *p) return true; } +bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion) +{ + switch ( io_completion ) + { + case HVMIO_realmode_completion: + { + struct hvm_emulate_ctxt ctxt; + + hvm_emulate_init_once(&ctxt, NULL, guest_cpu_user_regs()); + vmx_realmode_emulate_one(&ctxt); + hvm_emulate_writeback(&ctxt); + + break; + } + + default: + ASSERT_UNREACHABLE(); + break; + } + + return true; +} + bool handle_hvm_io_completion(struct vcpu *v) { struct domain *d = v->domain; @@ -209,19 +232,8 @@ bool handle_hvm_io_completion(struct vcpu *v) return handle_pio(vio->io_req.addr, vio->io_req.size, vio->io_req.dir); - case HVMIO_realmode_completion: - { - struct hvm_emulate_ctxt ctxt; - - hvm_emulate_init_once(&ctxt, NULL, guest_cpu_user_regs()); - vmx_realmode_emulate_one(&ctxt); - hvm_emulate_writeback(&ctxt); - - break; - } default: - ASSERT_UNREACHABLE(); - break; + return arch_vcpu_ioreq_completion(io_completion); } return true; @@ -477,9 +489,6 @@ static void hvm_update_ioreq_evtchn(struct hvm_ioreq_server *s, } } -#define HANDLE_BUFIOREQ(s) \ - ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) - static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, struct vcpu *v) { @@ -586,7 +595,7 @@ static void hvm_ioreq_server_remove_all_vcpus(struct hvm_ioreq_server *s) spin_unlock(&s->lock); } -static int hvm_ioreq_server_map_pages(struct hvm_ioreq_server *s) +int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s) { int rc; @@ -601,7 +610,7 @@ static int hvm_ioreq_server_map_pages(struct hvm_ioreq_server *s) return rc; } -static void hvm_ioreq_server_unmap_pages(struct hvm_ioreq_server *s) +void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s) { hvm_unmap_ioreq_gfn(s, true); hvm_unmap_ioreq_gfn(s, false); @@ -674,6 +683,12 @@ static int hvm_ioreq_server_alloc_rangesets(struct hvm_ioreq_server *s, return rc; } +void arch_ioreq_server_enable(struct hvm_ioreq_server *s) +{ + hvm_remove_ioreq_gfn(s, false); + hvm_remove_ioreq_gfn(s, true); +} + static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) { struct hvm_ioreq_vcpu *sv; @@ -683,8 +698,7 @@ static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) if ( s->enabled ) goto done; - hvm_remove_ioreq_gfn(s, false); - hvm_remove_ioreq_gfn(s, true); + arch_ioreq_server_enable(s); s->enabled = true; @@ -697,6 +711,12 @@ static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) spin_unlock(&s->lock); } +void arch_ioreq_server_disable(struct hvm_ioreq_server *s) +{ + hvm_add_ioreq_gfn(s, true); + hvm_add_ioreq_gfn(s, false); +} + static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) { spin_lock(&s->lock); @@ -704,8 +724,7 @@ static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) if ( !s->enabled ) goto done; - hvm_add_ioreq_gfn(s, true); - hvm_add_ioreq_gfn(s, false); + arch_ioreq_server_disable(s); s->enabled = false; @@ -750,7 +769,7 @@ static int hvm_ioreq_server_init(struct hvm_ioreq_server *s, fail_add: hvm_ioreq_server_remove_all_vcpus(s); - hvm_ioreq_server_unmap_pages(s); + arch_ioreq_server_unmap_pages(s); hvm_ioreq_server_free_rangesets(s); @@ -764,7 +783,7 @@ static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) hvm_ioreq_server_remove_all_vcpus(s); /* - * NOTE: It is safe to call both hvm_ioreq_server_unmap_pages() and + * NOTE: It is safe to call both arch_ioreq_server_unmap_pages() and * hvm_ioreq_server_free_pages() in that order. * This is because the former will do nothing if the pages * are not mapped, leaving the page to be freed by the latter. @@ -772,7 +791,7 @@ static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) * the page_info pointer to NULL, meaning the latter will do * nothing. */ - hvm_ioreq_server_unmap_pages(s); + arch_ioreq_server_unmap_pages(s); hvm_ioreq_server_free_pages(s); hvm_ioreq_server_free_rangesets(s); @@ -836,6 +855,12 @@ int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, return rc; } +/* Called when target domain is paused */ +void arch_ioreq_server_destroy(struct hvm_ioreq_server *s) +{ + p2m_set_ioreq_server(s->target, 0, s); +} + int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) { struct hvm_ioreq_server *s; @@ -855,7 +880,7 @@ int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) domain_pause(d); - p2m_set_ioreq_server(d, 0, s); + arch_ioreq_server_destroy(s); hvm_ioreq_server_disable(s); @@ -900,7 +925,7 @@ int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, if ( ioreq_gfn || bufioreq_gfn ) { - rc = hvm_ioreq_server_map_pages(s); + rc = arch_ioreq_server_map_pages(s); if ( rc ) goto out; } @@ -1080,6 +1105,22 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, return rc; } +/* Called with ioreq_server lock held */ +int arch_ioreq_server_map_mem_type(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags) +{ + return p2m_set_ioreq_server(d, flags, s); +} + +void arch_ioreq_server_map_mem_type_completed(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags) +{ + if ( flags == 0 && read_atomic(&p2m_get_hostp2m(d)->ioreq.entry_count) ) + p2m_change_entry_type_global(d, p2m_ioreq_server, p2m_ram_rw); +} + /* * Map or unmap an ioreq server to specific memory type. For now, only * HVMMEM_ioreq_server is supported, and in the future new types can be @@ -1112,18 +1153,13 @@ int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, if ( s->emulator != current->domain ) goto out; - rc = p2m_set_ioreq_server(d, flags, s); + rc = arch_ioreq_server_map_mem_type(d, s, flags); out: spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - if ( rc == 0 && flags == 0 ) - { - struct p2m_domain *p2m = p2m_get_hostp2m(d); - - if ( read_atomic(&p2m->ioreq.entry_count) ) - p2m_change_entry_type_global(d, p2m_ioreq_server, p2m_ram_rw); - } + if ( rc == 0 ) + arch_ioreq_server_map_mem_type_completed(d, s, flags); return rc; } @@ -1210,12 +1246,17 @@ void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); } +bool arch_ioreq_server_destroy_all(struct domain *d) +{ + return relocate_portio_handler(d, 0xcf8, 0xcf8, 4); +} + void hvm_destroy_all_ioreq_servers(struct domain *d) { struct hvm_ioreq_server *s; unsigned int id; - if ( !relocate_portio_handler(d, 0xcf8, 0xcf8, 4) ) + if ( !arch_ioreq_server_destroy_all(d) ) return; spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); @@ -1239,33 +1280,28 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); } -struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p) +bool arch_ioreq_server_get_type_addr(const struct domain *d, + const ioreq_t *p, + uint8_t *type, + uint64_t *addr) { - struct hvm_ioreq_server *s; - uint32_t cf8; - uint8_t type; - uint64_t addr; - unsigned int id; + unsigned int cf8 = d->arch.hvm.pci_cf8; if ( p->type != IOREQ_TYPE_COPY && p->type != IOREQ_TYPE_PIO ) - return NULL; - - cf8 = d->arch.hvm.pci_cf8; + return false; if ( p->type == IOREQ_TYPE_PIO && (p->addr & ~3) == 0xcfc && CF8_ENABLED(cf8) ) { - uint32_t x86_fam; + unsigned int x86_fam, reg; pci_sbdf_t sbdf; - unsigned int reg; reg = hvm_pci_decode_addr(cf8, p->addr, &sbdf); /* PCI config data cycle */ - type = XEN_DMOP_IO_RANGE_PCI; - addr = ((uint64_t)sbdf.sbdf << 32) | reg; + *type = XEN_DMOP_IO_RANGE_PCI; + *addr = ((uint64_t)sbdf.sbdf << 32) | reg; /* AMD extended configuration space access? */ if ( CF8_ADDR_HI(cf8) && d->arch.cpuid->x86_vendor == X86_VENDOR_AMD && @@ -1277,16 +1313,30 @@ struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, if ( !rdmsr_safe(MSR_AMD64_NB_CFG, msr_val) && (msr_val & (1ULL << AMD64_NB_CFG_CF8_EXT_ENABLE_BIT)) ) - addr |= CF8_ADDR_HI(cf8); + *addr |= CF8_ADDR_HI(cf8); } } else { - type = (p->type == IOREQ_TYPE_PIO) ? - XEN_DMOP_IO_RANGE_PORT : XEN_DMOP_IO_RANGE_MEMORY; - addr = p->addr; + *type = (p->type == IOREQ_TYPE_PIO) ? + XEN_DMOP_IO_RANGE_PORT : XEN_DMOP_IO_RANGE_MEMORY; + *addr = p->addr; } + return true; +} + +struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, + ioreq_t *p) +{ + struct hvm_ioreq_server *s; + uint8_t type; + uint64_t addr; + unsigned int id; + + if ( !arch_ioreq_server_get_type_addr(d, p, &type, &addr) ) + return NULL; + FOR_EACH_IOREQ_SERVER(d, id, s) { struct rangeset *r; @@ -1515,11 +1565,16 @@ static int hvm_access_cf8( return X86EMUL_UNHANDLEABLE; } +void arch_ioreq_domain_init(struct domain *d) +{ + register_portio_handler(d, 0xcf8, 4, hvm_access_cf8); +} + void hvm_ioreq_init(struct domain *d) { spin_lock_init(&d->arch.hvm.ioreq_server.lock); - register_portio_handler(d, 0xcf8, 4, hvm_access_cf8); + arch_ioreq_domain_init(d); } /* diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h new file mode 100644 index 0000000..d0980c5 --- /dev/null +++ b/xen/include/xen/ioreq.h @@ -0,0 +1,54 @@ +/* + * ioreq.h: Hardware virtual machine assist interface definitions. + * + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#ifndef __XEN_IOREQ_H__ +#define __XEN_IOREQ_H__ + +#include + +#define HANDLE_BUFIOREQ(s) \ + ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) + +bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); +int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s); +void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s); +void arch_ioreq_server_enable(struct hvm_ioreq_server *s); +void arch_ioreq_server_disable(struct hvm_ioreq_server *s); +void arch_ioreq_server_destroy(struct hvm_ioreq_server *s); +int arch_ioreq_server_map_mem_type(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags); +void arch_ioreq_server_map_mem_type_completed(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags); +bool arch_ioreq_server_destroy_all(struct domain *d); +bool arch_ioreq_server_get_type_addr(const struct domain *d, const ioreq_t *p, + uint8_t *type, uint64_t *addr); +void arch_ioreq_domain_init(struct domain *d); + +#endif /* __XEN_IOREQ_H__ */ + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ From patchwork Mon Jan 25 19:08:09 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044031 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D0644C433E0 for ; Mon, 25 Jan 2021 19:09:09 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 6EB5B21D79 for ; Mon, 25 Jan 2021 19:09:09 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 6EB5B21D79 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74297.133513 (Exim 4.92) (envelope-from ) id 1l47ET-0008Px-Po; Mon, 25 Jan 2021 19:09:01 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74297.133513; Mon, 25 Jan 2021 19:09:01 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47ET-0008Pq-LR; Mon, 25 Jan 2021 19:09:01 +0000 Received: by outflank-mailman (input) for mailman id 74297; Mon, 25 Jan 2021 19:09:00 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47ES-0008N7-Ff for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:00 +0000 Received: from mail-wr1-x430.google.com (unknown [2a00:1450:4864:20::430]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id a5de0fab-65d0-4a32-8bdd-173fcf92abf3; Mon, 25 Jan 2021 19:08:55 +0000 (UTC) Received: by mail-wr1-x430.google.com with SMTP id g10so14126339wrx.1 for ; Mon, 25 Jan 2021 11:08:55 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.53 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:53 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: a5de0fab-65d0-4a32-8bdd-173fcf92abf3 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=OmZ68sThbF2Pr1tzxlriA0ATbzYpV+ZK7uDIZSpg8u0=; b=n/XDYAY3Odji1o91Ht2hy3VfnEDogmW13d3mivdQRhq9LwwlT+NNC3rctots2K5IUD Hig1ZLM5ddNxQJrHDBkXRo410Flanld81UvRFev5h5WBJ6k1gs3yRfq6YaZF8SOod4qk 9Hjq2aV8MNG+Hvk3Ihtrxw2r7IvdH11x9UlZJddNQJb3Sc068tutdrsyEry0aA0eTt83 8i7MyanlCthGsJ06nTaYTRwGwBzdQPjHg/uIYW0EWvFyEwnL9tEQAOjPrakWVJqzGhnb myOgwQRLJG54ZD/kXpwlqKHUQamiiSuBpMQ1dFD1AnZhicsA40CLkLS9B4TBeXIvY1Hk tIxw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=OmZ68sThbF2Pr1tzxlriA0ATbzYpV+ZK7uDIZSpg8u0=; b=GSmXLf4UVFbRYyoB++exZyr/oA32T5/BlgusmmINwuN2q0pHSy2vgoDVQ4+v0R0n9K s8h2iKKC3jWV6RVyTNZNCBGuB67/mBsxA1wH1EVWgCOCT0t2Qez7WGWw8hMoY+J4YmBg 5Anvyy0DX4bWviJr23R9RwmEXLjLHNowWZ7BWehvk8ibKqRo0NiQZG/CjIaNO7tMZreb UAS95wATS5hdttZXPoT8b/VEPMNcpHTaHMgzfcF2iVfs/Q/CeANuekO4idtvzJmt1oWH jC61EjXi5DuYMeP5mqJcn4zSqjRjLtRKecEJB//DzoKxtRP7TwfRyI8Egl/P1KIYh9nT mfTA== X-Gm-Message-State: AOAM5333+I1+tNlFov16EpnbEvkUcTL8rXFdSiRvLs8fnSb7zp46Y2id 3aGOsS007fhHUHRG53yaT/zV3CmXDvfWDw== X-Google-Smtp-Source: ABdhPJyHFB6Lt2TWXqzfTdOccPKyoaKm+Vm5ffKwBCgxnIF7x2DXPF1Wvd2vBY/fW4W4ulc4PrKv5w== X-Received: by 2002:a5d:65ca:: with SMTP id e10mr2617075wrw.166.1611601734390; Mon, 25 Jan 2021 11:08:54 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 02/22] x86/ioreq: Add IOREQ_STATUS_* #define-s and update code for moving Date: Mon, 25 Jan 2021 21:08:09 +0200 Message-Id: <1611601709-28361-3-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko This patch continues to make some preparation to x86/hvm/ioreq.c before moving to the common code. Add IOREQ_STATUS_* #define-s and update candidates for moving since X86EMUL_* shouldn't be exposed to the common code in that form. This support is going to be used on Arm to be able run device emulator outside of Xen hypervisor. Signed-off-by: Oleksandr Tyshchenko Acked-by: Jan Beulich Reviewed-by: Alex Bennée Reviewed-by: Julien Grall Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V2 -> V3: - new patch, was split from [PATCH V2 01/23] x86/ioreq: Prepare IOREQ feature for making it common Changes V3 -> V4: - add Alex's R-b and Jan's A-b - add a comment above IOREQ_STATUS_* #define-s Changes V4 -> V5: - rebase - add Julien's and Paul's R-b --- --- xen/arch/x86/hvm/ioreq.c | 16 ++++++++-------- xen/include/asm-x86/hvm/ioreq.h | 5 +++++ 2 files changed, 13 insertions(+), 8 deletions(-) diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 3c3c173..27a4a6f 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -1401,7 +1401,7 @@ static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) pg = iorp->va; if ( !pg ) - return X86EMUL_UNHANDLEABLE; + return IOREQ_STATUS_UNHANDLED; /* * Return 0 for the cases we can't deal with: @@ -1431,7 +1431,7 @@ static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) break; default: gdprintk(XENLOG_WARNING, "unexpected ioreq size: %u\n", p->size); - return X86EMUL_UNHANDLEABLE; + return IOREQ_STATUS_UNHANDLED; } spin_lock(&s->bufioreq_lock); @@ -1441,7 +1441,7 @@ static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) { /* The queue is full: send the iopacket through the normal path. */ spin_unlock(&s->bufioreq_lock); - return X86EMUL_UNHANDLEABLE; + return IOREQ_STATUS_UNHANDLED; } pg->buf_ioreq[pg->ptrs.write_pointer % IOREQ_BUFFER_SLOT_NUM] = bp; @@ -1472,7 +1472,7 @@ static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) notify_via_xen_event_channel(d, s->bufioreq_evtchn); spin_unlock(&s->bufioreq_lock); - return X86EMUL_OKAY; + return IOREQ_STATUS_HANDLED; } int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, @@ -1488,7 +1488,7 @@ int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, return hvm_send_buffered_ioreq(s, proto_p); if ( unlikely(!vcpu_start_shutdown_deferral(curr)) ) - return X86EMUL_RETRY; + return IOREQ_STATUS_RETRY; list_for_each_entry ( sv, &s->ioreq_vcpu_list, @@ -1528,11 +1528,11 @@ int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, notify_via_xen_event_channel(d, port); sv->pending = true; - return X86EMUL_RETRY; + return IOREQ_STATUS_RETRY; } } - return X86EMUL_UNHANDLEABLE; + return IOREQ_STATUS_UNHANDLED; } unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) @@ -1546,7 +1546,7 @@ unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) if ( !s->enabled ) continue; - if ( hvm_send_ioreq(s, p, buffered) == X86EMUL_UNHANDLEABLE ) + if ( hvm_send_ioreq(s, p, buffered) == IOREQ_STATUS_UNHANDLED ) failed++; } diff --git a/xen/include/asm-x86/hvm/ioreq.h b/xen/include/asm-x86/hvm/ioreq.h index e2588e9..df0c292 100644 --- a/xen/include/asm-x86/hvm/ioreq.h +++ b/xen/include/asm-x86/hvm/ioreq.h @@ -55,6 +55,11 @@ unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); void hvm_ioreq_init(struct domain *d); +/* This correlation must not be altered */ +#define IOREQ_STATUS_HANDLED X86EMUL_OKAY +#define IOREQ_STATUS_UNHANDLED X86EMUL_UNHANDLEABLE +#define IOREQ_STATUS_RETRY X86EMUL_RETRY + #endif /* __ASM_X86_HVM_IOREQ_H__ */ /* From patchwork Mon Jan 25 19:08:10 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044037 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id F0540C433DB for ; Mon, 25 Jan 2021 19:09:21 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id AE52D216FD for ; Mon, 25 Jan 2021 19:09:21 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org AE52D216FD Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74299.133537 (Exim 4.92) (envelope-from ) id 1l47Ee-00008P-EF; Mon, 25 Jan 2021 19:09:12 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74299.133537; Mon, 25 Jan 2021 19:09:12 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ee-00008G-Ac; Mon, 25 Jan 2021 19:09:12 +0000 Received: by outflank-mailman (input) for mailman id 74299; Mon, 25 Jan 2021 19:09:10 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ec-0008N7-GC for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:10 +0000 Received: from mail-wm1-x32c.google.com (unknown [2a00:1450:4864:20::32c]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id a001e2ed-890d-44c7-a0c3-1544d1748298; Mon, 25 Jan 2021 19:08:56 +0000 (UTC) Received: by mail-wm1-x32c.google.com with SMTP id m187so387134wme.2 for ; Mon, 25 Jan 2021 11:08:56 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.54 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:54 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: a001e2ed-890d-44c7-a0c3-1544d1748298 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=OsLFEZreg7YqeEEhgK+SjAQbSMvohSeO9g6je9KBEsI=; b=T9dfsa7gmJ7W9ij7E90dOw8sQ/oEP/vrOLg+A0tjx3SjZX4SrPb2qHgRYjl0PEz4Kk rk7w2UJWqOUt/sUs8II+69Lyx4U5q/TgT3A9isaOt4HDH8XvXgVZE8LWQYEHT/kjA9Y/ sGc0/fc6a4BPVVORdbe7QgkGTNbU7K69s19JY/LIxU1TC2W4YasNYnJILuy7XcfdFA/t k18laMJGxy+Rfd/iXdogqLPOyge+FU+fWAd6TpRcAsh2ALXeaLhDDLXtVt6OXOEJ62q4 yAxveXM2N0oVD+c/yemTrUIUSJz2dTKdFsgVaur6CH6YVjHfCqzjhtEpwpr5XgO9ZBXF dbJA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=OsLFEZreg7YqeEEhgK+SjAQbSMvohSeO9g6je9KBEsI=; b=spEGlAgaZxCOjcV89sijeFe7qmeShWouMkvL9HB64lq9++5RkAdzTO62wLq9CVCD1F vVtXoc1EvdgDvKuUs1pIb4YBDKNtCd6K//Ivx6QIcYQzEv3Y/4PjcQmitLZZv458xjmp VLIHFoLAy5Htb6Jv4vsgmsapt/4p71XT7ElVFq07T1gnn5AaopOUNNm/fj9MrRJLa/4d jAdtj3+6/J49hbMB2p5Q3KKhLCBs7XFBw5s1HTrzFll+L8ZcYK7pWQfJ39kBpU58ZgVj 7EdQTu+21mIocOuuXto/7jPNsND24/6fI6yjOAwelqjvCbZVvtbeMdlkwFdUW+nDrjYV ME9A== X-Gm-Message-State: AOAM531kRpOWOYMGGaVDkotLwOOdtvXRS1NAUzDLWeprhfG1gK4uQSqA x0UgjeqBzATNr2beYUAqECyjIXJ7Q2wdLg== X-Google-Smtp-Source: ABdhPJwicXK5X8XYRb1eHofoVzV3U5TxlFZM1E98xbSQrWkgmoNYx3frsyt+PgqQkKp03Nm5JT78/g== X-Received: by 2002:a7b:c09a:: with SMTP id r26mr1494547wmh.64.1611601735366; Mon, 25 Jan 2021 11:08:55 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 03/22] x86/ioreq: Provide out-of-line wrapper for the handle_mmio() Date: Mon, 25 Jan 2021 21:08:10 +0200 Message-Id: <1611601709-28361-4-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko The IOREQ is about to be common feature and Arm will have its own implementation. But the name of the function is pretty generic and can be confusing on Arm (we already have a try_handle_mmio()). In order not to rename the function (which is used for a varying set of purposes on x86) globally and get non-confusing variant on Arm provide a wrapper arch_ioreq_complete_mmio() to be used on common and Arm code. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Jan Beulich Reviewed-by: Alex Bennée Reviewed-by: Julien Grall Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch Changes V1 -> V2: - remove "handle" - add Jan's A-b Changes V2 -> V3: - remove Jan's A-b - update patch subject/description - use out-of-line function instead of #define - put earlier in the series to avoid breakage Changes V3 -> V4: - add Jan's R-b - rename ioreq_complete_mmio() to arch_ioreq_complete_mmio() Changes V4 -> V5: - rebase - add Alex's, Julien's and Paul's R-b --- --- xen/arch/x86/hvm/ioreq.c | 7 ++++++- xen/include/xen/ioreq.h | 1 + 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 27a4a6f..30e8724 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -36,6 +36,11 @@ #include #include +bool arch_ioreq_complete_mmio(void) +{ + return handle_mmio(); +} + static void set_ioreq_server(struct domain *d, unsigned int id, struct hvm_ioreq_server *s) { @@ -226,7 +231,7 @@ bool handle_hvm_io_completion(struct vcpu *v) break; case HVMIO_mmio_completion: - return handle_mmio(); + return arch_ioreq_complete_mmio(); case HVMIO_pio_completion: return handle_pio(vio->io_req.addr, vio->io_req.size, diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index d0980c5..b95d3ef 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -24,6 +24,7 @@ #define HANDLE_BUFIOREQ(s) \ ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) +bool arch_ioreq_complete_mmio(void); bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s); void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s); From patchwork Mon Jan 25 19:08:11 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044045 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id AA031C433DB for ; Mon, 25 Jan 2021 19:09:38 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 59DA220E65 for ; Mon, 25 Jan 2021 19:09:38 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 59DA220E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74303.133572 (Exim 4.92) (envelope-from ) id 1l47Et-0000P3-Ha; Mon, 25 Jan 2021 19:09:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74303.133572; Mon, 25 Jan 2021 19:09:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Et-0000Os-E8; Mon, 25 Jan 2021 19:09:27 +0000 Received: by outflank-mailman (input) for mailman id 74303; Mon, 25 Jan 2021 19:09:25 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Er-0008N7-GT for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:25 +0000 Received: from mail-wm1-x332.google.com (unknown [2a00:1450:4864:20::332]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 1a4481d6-c3dd-413a-9ead-823f9c9f5a49; Mon, 25 Jan 2021 19:08:58 +0000 (UTC) Received: by mail-wm1-x332.google.com with SMTP id m2so398095wmm.1 for ; Mon, 25 Jan 2021 11:08:58 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.55 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:56 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 1a4481d6-c3dd-413a-9ead-823f9c9f5a49 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=ou2s10ByENZ3UWEosFFMxi35IpzLY08hMuYeGXauNao=; b=PYbQubDyli3U3WACP96E0twNrMv/oPfsg4WtO0WxD8/jIKqNDhumQvw/be2GuxF43m fUeXvD/JzRbY4rZvVALgZsGRXQyaYCPY3qfmc/DtH9S2nHtKzqhHp73oAUNL4vlC46eD LwcKrbuQoejW1SNBZ/cosnSLhC/KU4Iojk+JMr30QUNlNO9FpdbFHEpC4QMzCHjZyLIu RRcWezP7DBQ8eM5e3a3jhlSsqhQCDLtCnsBMrv9mJSU5igOCVnfPHbFs6ku9hnd4406j H3Q77e5E8PmaR+KsWUNl01BWFBAJPMMjWNIQ1URDBnW803t3CwTQbCv/Kb3Nk0aECs7c Etpg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=ou2s10ByENZ3UWEosFFMxi35IpzLY08hMuYeGXauNao=; b=Uj5wSoJx+F1UTW5o5iV31fbcS2YLu4rfAkMIHAuUQfgXSGdZCBM2wf6rMdrB1yzduq 2Ib8eoY3oPAeKXY/72OnQKtRWBHoY2TR/qII8YFMdZQ0JlmTYEWmcIy9eDGML0CcUBpx C7pUmSpN6IRDg5cfDzJxp+8kyqJT5NSW52/Ps9zWOUo/zxyb8WyNa9vUo8mWBPtkyaTz vSIssHwEnNyNugOBIGiJxLjJ211neg2JkaSBuZOAAEnhkCmeRxqxNyaum3A52Qm6DEfJ 4ZUzjKxlip9qYajuLjD5eBWlgB/2myVoYuswgub5Z5YJuZbh2CkGAkCGKT2MFcUG+I9n ofGQ== X-Gm-Message-State: AOAM5323kPJeWejpthkuDmmB6Vm7RgJrtEuF3gPhmfxoa1rc6ad6gb6s EaUUs0ucAMhc8KsMpT5bCImCG9rVZNb7xw== X-Google-Smtp-Source: ABdhPJwWVcOHZbY99ahc9Tt/f/1JQ/t5X+D2LJnMbHcztQVwX6gWu5T6Qlm/kvYzkydSvKPzwZa8Ig== X-Received: by 2002:a05:600c:154f:: with SMTP id f15mr1501404wmg.46.1611601736728; Mon, 25 Jan 2021 11:08:56 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Andrew Cooper , George Dunlap , Ian Jackson , Jan Beulich , Julien Grall , Stefano Stabellini , Wei Liu , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Paul Durrant , Jun Nakajima , Kevin Tian , Tim Deegan , Julien Grall Subject: [PATCH V5 04/22] xen/ioreq: Make x86's IOREQ feature common Date: Mon, 25 Jan 2021 21:08:11 +0200 Message-Id: <1611601709-28361-5-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko As a lot of x86 code can be re-used on Arm later on, this patch moves previously prepared IOREQ support to the common code (the code movement is verbatim copy). The "legacy" mechanism of mapping magic pages for the IOREQ servers remains x86 specific and not exposed to the common code. The common IOREQ feature is supposed to be built with IOREQ_SERVER option enabled, which is selected for x86's config HVM for now. In order to avoid having a gigantic patch here, the subsequent patches will update remaining bits in the common code step by step: - Make IOREQ related structs/materials common - Drop the "hvm" prefixes and infixes - Remove layering violation by moving corresponding fields out of *arch.hvm* or abstracting away accesses to them Introduce asm/ioreq.h wrapper to be included by common ioreq.h instead of asm/hvm/ioreq.h to avoid HVM-ism in the code common. Also include which will be needed on Arm to avoid touch the common code again when introducing Arm specific bits. This support is going to be used on Arm to be able run device emulator outside of Xen hypervisor. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen Acked-by: Jan Beulich --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" *** Please note, this patch depends on the following which is on review: https://patchwork.kernel.org/patch/11816689/ The effort (to get it upstreamed) was paused because of the security issue around that code (XSA-348). *** Changes RFC -> V1: - was split into three patches: - x86/ioreq: Prepare IOREQ feature for making it common - xen/ioreq: Make x86's IOREQ feature common - xen/ioreq: Make x86's hvm_ioreq_needs_completion() common - update MAINTAINERS file - do not use a separate subdir for the IOREQ stuff, move it to: - xen/common/ioreq.c - xen/include/xen/ioreq.h - update x86's files to include xen/ioreq.h - remove unneeded headers in arch/x86/hvm/ioreq.c - re-order the headers alphabetically in common/ioreq.c - update common/ioreq.c according to the newly introduced arch functions: arch_hvm_destroy_ioreq_server()/arch_handle_hvm_io_completion() Changes V1 -> V2: - update patch description - make everything needed in the previous patch to achieve a truly rename here - don't include unnecessary headers from asm-x86/hvm/ioreq.h and xen/ioreq.h - use __XEN_IOREQ_H__ instead of __IOREQ_H__ - move get_ioreq_server() to common/ioreq.c Changes V2 -> V3: - update patch description - make everything needed in the previous patch to not expose "legacy" interface to the common code here - update patch according the "legacy interface" is x86 specific - include in common ioreq.c Changes V3 -> V4: - rebase - don't include from arch header - мove all arch hook declarations to the common header Change V4 -> V5: - rebase - introduce asm-x86/ioreq.h wrapper: - update MAINTAINERS file - include asm/ioreq.h instead of asm/hvm/ioreq.h from common ioreq.c - include public/hvm/dm_op.h from common ioreq.h - add Paul's R-b --- --- MAINTAINERS | 9 +- xen/arch/x86/Kconfig | 1 + xen/arch/x86/hvm/dm.c | 2 +- xen/arch/x86/hvm/emulate.c | 2 +- xen/arch/x86/hvm/hvm.c | 2 +- xen/arch/x86/hvm/io.c | 2 +- xen/arch/x86/hvm/ioreq.c | 1316 ++------------------------------------- xen/arch/x86/hvm/stdvga.c | 2 +- xen/arch/x86/hvm/vmx/vvmx.c | 3 +- xen/arch/x86/mm.c | 2 +- xen/arch/x86/mm/shadow/common.c | 2 +- xen/common/Kconfig | 3 + xen/common/Makefile | 1 + xen/common/ioreq.c | 1290 ++++++++++++++++++++++++++++++++++++++ xen/include/asm-x86/hvm/ioreq.h | 36 -- xen/include/asm-x86/ioreq.h | 37 ++ xen/include/xen/ioreq.h | 38 ++ 17 files changed, 1422 insertions(+), 1326 deletions(-) create mode 100644 xen/common/ioreq.c create mode 100644 xen/include/asm-x86/ioreq.h diff --git a/MAINTAINERS b/MAINTAINERS index 5079b83..c4f9aff 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -333,6 +333,13 @@ X: xen/drivers/passthrough/vtd/ X: xen/drivers/passthrough/device_tree.c F: xen/include/xen/iommu.h +I/O EMULATION (IOREQ) +M: Paul Durrant +S: Supported +F: xen/common/ioreq.c +F: xen/include/xen/ioreq.h +F: xen/include/public/hvm/ioreq.h + KCONFIG M: Doug Goldstein S: Supported @@ -549,7 +556,7 @@ F: xen/arch/x86/hvm/ioreq.c F: xen/include/asm-x86/hvm/emulate.h F: xen/include/asm-x86/hvm/io.h F: xen/include/asm-x86/hvm/ioreq.h -F: xen/include/public/hvm/ioreq.h +F: xen/include/asm-x86/ioreq.h X86 MEMORY MANAGEMENT M: Jan Beulich diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig index 78f351f..ea9a9ea 100644 --- a/xen/arch/x86/Kconfig +++ b/xen/arch/x86/Kconfig @@ -92,6 +92,7 @@ config PV_LINEAR_PT config HVM def_bool !PV_SHIM_EXCLUSIVE + select IOREQ_SERVER prompt "HVM support" ---help--- Interfaces to support HVM domains. HVM domains require hardware diff --git a/xen/arch/x86/hvm/dm.c b/xen/arch/x86/hvm/dm.c index 71f5ca4..d3e2a9e 100644 --- a/xen/arch/x86/hvm/dm.c +++ b/xen/arch/x86/hvm/dm.c @@ -17,12 +17,12 @@ #include #include #include +#include #include #include #include #include -#include #include #include diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c index 24cf85f..60ca465 100644 --- a/xen/arch/x86/hvm/emulate.c +++ b/xen/arch/x86/hvm/emulate.c @@ -10,6 +10,7 @@ */ #include +#include #include #include #include @@ -20,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c index 54e32e4..bc96947 100644 --- a/xen/arch/x86/hvm/hvm.c +++ b/xen/arch/x86/hvm/hvm.c @@ -20,6 +20,7 @@ #include #include +#include #include #include #include @@ -64,7 +65,6 @@ #include #include #include -#include #include #include #include diff --git a/xen/arch/x86/hvm/io.c b/xen/arch/x86/hvm/io.c index 3e09d9b..11e007d 100644 --- a/xen/arch/x86/hvm/io.c +++ b/xen/arch/x86/hvm/io.c @@ -19,6 +19,7 @@ */ #include +#include #include #include #include @@ -35,7 +36,6 @@ #include #include #include -#include #include #include #include diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 30e8724..666d695 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -30,7 +30,6 @@ #include #include -#include #include #include @@ -41,140 +40,6 @@ bool arch_ioreq_complete_mmio(void) return handle_mmio(); } -static void set_ioreq_server(struct domain *d, unsigned int id, - struct hvm_ioreq_server *s) -{ - ASSERT(id < MAX_NR_IOREQ_SERVERS); - ASSERT(!s || !d->arch.hvm.ioreq_server.server[id]); - - d->arch.hvm.ioreq_server.server[id] = s; -} - -#define GET_IOREQ_SERVER(d, id) \ - (d)->arch.hvm.ioreq_server.server[id] - -static struct hvm_ioreq_server *get_ioreq_server(const struct domain *d, - unsigned int id) -{ - if ( id >= MAX_NR_IOREQ_SERVERS ) - return NULL; - - return GET_IOREQ_SERVER(d, id); -} - -/* - * Iterate over all possible ioreq servers. - * - * NOTE: The iteration is backwards such that more recently created - * ioreq servers are favoured in hvm_select_ioreq_server(). - * This is a semantic that previously existed when ioreq servers - * were held in a linked list. - */ -#define FOR_EACH_IOREQ_SERVER(d, id, s) \ - for ( (id) = MAX_NR_IOREQ_SERVERS; (id) != 0; ) \ - if ( !(s = GET_IOREQ_SERVER(d, --(id))) ) \ - continue; \ - else - -static ioreq_t *get_ioreq(struct hvm_ioreq_server *s, struct vcpu *v) -{ - shared_iopage_t *p = s->ioreq.va; - - ASSERT((v == current) || !vcpu_runnable(v)); - ASSERT(p != NULL); - - return &p->vcpu_ioreq[v->vcpu_id]; -} - -static struct hvm_ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, - struct hvm_ioreq_server **srvp) -{ - struct domain *d = v->domain; - struct hvm_ioreq_server *s; - unsigned int id; - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - struct hvm_ioreq_vcpu *sv; - - list_for_each_entry ( sv, - &s->ioreq_vcpu_list, - list_entry ) - { - if ( sv->vcpu == v && sv->pending ) - { - if ( srvp ) - *srvp = s; - return sv; - } - } - } - - return NULL; -} - -bool hvm_io_pending(struct vcpu *v) -{ - return get_pending_vcpu(v, NULL); -} - -static bool hvm_wait_for_io(struct hvm_ioreq_vcpu *sv, ioreq_t *p) -{ - unsigned int prev_state = STATE_IOREQ_NONE; - unsigned int state = p->state; - uint64_t data = ~0; - - smp_rmb(); - - /* - * The only reason we should see this condition be false is when an - * emulator dying races with I/O being requested. - */ - while ( likely(state != STATE_IOREQ_NONE) ) - { - if ( unlikely(state < prev_state) ) - { - gdprintk(XENLOG_ERR, "Weird HVM ioreq state transition %u -> %u\n", - prev_state, state); - sv->pending = false; - domain_crash(sv->vcpu->domain); - return false; /* bail */ - } - - switch ( prev_state = state ) - { - case STATE_IORESP_READY: /* IORESP_READY -> NONE */ - p->state = STATE_IOREQ_NONE; - data = p->data; - break; - - case STATE_IOREQ_READY: /* IOREQ_{READY,INPROCESS} -> IORESP_READY */ - case STATE_IOREQ_INPROCESS: - wait_on_xen_event_channel(sv->ioreq_evtchn, - ({ state = p->state; - smp_rmb(); - state != prev_state; })); - continue; - - default: - gdprintk(XENLOG_ERR, "Weird HVM iorequest state %u\n", state); - sv->pending = false; - domain_crash(sv->vcpu->domain); - return false; /* bail */ - } - - break; - } - - p = &sv->vcpu->arch.hvm.hvm_io.io_req; - if ( hvm_ioreq_needs_completion(p) ) - p->data = data; - - sv->pending = false; - - return true; -} - bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion) { switch ( io_completion ) @@ -198,52 +63,6 @@ bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion) return true; } -bool handle_hvm_io_completion(struct vcpu *v) -{ - struct domain *d = v->domain; - struct hvm_vcpu_io *vio = &v->arch.hvm.hvm_io; - struct hvm_ioreq_server *s; - struct hvm_ioreq_vcpu *sv; - enum hvm_io_completion io_completion; - - if ( has_vpci(d) && vpci_process_pending(v) ) - { - raise_softirq(SCHEDULE_SOFTIRQ); - return false; - } - - sv = get_pending_vcpu(v, &s); - if ( sv && !hvm_wait_for_io(sv, get_ioreq(s, v)) ) - return false; - - vio->io_req.state = hvm_ioreq_needs_completion(&vio->io_req) ? - STATE_IORESP_READY : STATE_IOREQ_NONE; - - msix_write_completion(v); - vcpu_end_shutdown_deferral(v); - - io_completion = vio->io_completion; - vio->io_completion = HVMIO_no_completion; - - switch ( io_completion ) - { - case HVMIO_no_completion: - break; - - case HVMIO_mmio_completion: - return arch_ioreq_complete_mmio(); - - case HVMIO_pio_completion: - return handle_pio(vio->io_req.addr, vio->io_req.size, - vio->io_req.dir); - - default: - return arch_vcpu_ioreq_completion(io_completion); - } - - return true; -} - static gfn_t hvm_alloc_legacy_ioreq_gfn(struct hvm_ioreq_server *s) { struct domain *d = s->target; @@ -360,93 +179,6 @@ static int hvm_map_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) return rc; } -static int hvm_alloc_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) -{ - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; - struct page_info *page; - - if ( iorp->page ) - { - /* - * If a guest frame has already been mapped (which may happen - * on demand if hvm_get_ioreq_server_info() is called), then - * allocating a page is not permitted. - */ - if ( !gfn_eq(iorp->gfn, INVALID_GFN) ) - return -EPERM; - - return 0; - } - - page = alloc_domheap_page(s->target, MEMF_no_refcount); - - if ( !page ) - return -ENOMEM; - - if ( !get_page_and_type(page, s->target, PGT_writable_page) ) - { - /* - * The domain can't possibly know about this page yet, so failure - * here is a clear indication of something fishy going on. - */ - domain_crash(s->emulator); - return -ENODATA; - } - - iorp->va = __map_domain_page_global(page); - if ( !iorp->va ) - goto fail; - - iorp->page = page; - clear_page(iorp->va); - return 0; - - fail: - put_page_alloc_ref(page); - put_page_and_type(page); - - return -ENOMEM; -} - -static void hvm_free_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) -{ - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; - struct page_info *page = iorp->page; - - if ( !page ) - return; - - iorp->page = NULL; - - unmap_domain_page_global(iorp->va); - iorp->va = NULL; - - put_page_alloc_ref(page); - put_page_and_type(page); -} - -bool is_ioreq_server_page(struct domain *d, const struct page_info *page) -{ - const struct hvm_ioreq_server *s; - unsigned int id; - bool found = false; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - if ( (s->ioreq.page == page) || (s->bufioreq.page == page) ) - { - found = true; - break; - } - } - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return found; -} - static void hvm_remove_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) { @@ -481,125 +213,6 @@ static int hvm_add_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) return rc; } -static void hvm_update_ioreq_evtchn(struct hvm_ioreq_server *s, - struct hvm_ioreq_vcpu *sv) -{ - ASSERT(spin_is_locked(&s->lock)); - - if ( s->ioreq.va != NULL ) - { - ioreq_t *p = get_ioreq(s, sv->vcpu); - - p->vp_eport = sv->ioreq_evtchn; - } -} - -static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, - struct vcpu *v) -{ - struct hvm_ioreq_vcpu *sv; - int rc; - - sv = xzalloc(struct hvm_ioreq_vcpu); - - rc = -ENOMEM; - if ( !sv ) - goto fail1; - - spin_lock(&s->lock); - - rc = alloc_unbound_xen_event_channel(v->domain, v->vcpu_id, - s->emulator->domain_id, NULL); - if ( rc < 0 ) - goto fail2; - - sv->ioreq_evtchn = rc; - - if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) - { - rc = alloc_unbound_xen_event_channel(v->domain, 0, - s->emulator->domain_id, NULL); - if ( rc < 0 ) - goto fail3; - - s->bufioreq_evtchn = rc; - } - - sv->vcpu = v; - - list_add(&sv->list_entry, &s->ioreq_vcpu_list); - - if ( s->enabled ) - hvm_update_ioreq_evtchn(s, sv); - - spin_unlock(&s->lock); - return 0; - - fail3: - free_xen_event_channel(v->domain, sv->ioreq_evtchn); - - fail2: - spin_unlock(&s->lock); - xfree(sv); - - fail1: - return rc; -} - -static void hvm_ioreq_server_remove_vcpu(struct hvm_ioreq_server *s, - struct vcpu *v) -{ - struct hvm_ioreq_vcpu *sv; - - spin_lock(&s->lock); - - list_for_each_entry ( sv, - &s->ioreq_vcpu_list, - list_entry ) - { - if ( sv->vcpu != v ) - continue; - - list_del(&sv->list_entry); - - if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) - free_xen_event_channel(v->domain, s->bufioreq_evtchn); - - free_xen_event_channel(v->domain, sv->ioreq_evtchn); - - xfree(sv); - break; - } - - spin_unlock(&s->lock); -} - -static void hvm_ioreq_server_remove_all_vcpus(struct hvm_ioreq_server *s) -{ - struct hvm_ioreq_vcpu *sv, *next; - - spin_lock(&s->lock); - - list_for_each_entry_safe ( sv, - next, - &s->ioreq_vcpu_list, - list_entry ) - { - struct vcpu *v = sv->vcpu; - - list_del(&sv->list_entry); - - if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) - free_xen_event_channel(v->domain, s->bufioreq_evtchn); - - free_xen_event_channel(v->domain, sv->ioreq_evtchn); - - xfree(sv); - } - - spin_unlock(&s->lock); -} - int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s) { int rc; @@ -621,688 +234,63 @@ void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s) hvm_unmap_ioreq_gfn(s, false); } -static int hvm_ioreq_server_alloc_pages(struct hvm_ioreq_server *s) -{ - int rc; - - rc = hvm_alloc_ioreq_mfn(s, false); - - if ( !rc && (s->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) ) - rc = hvm_alloc_ioreq_mfn(s, true); - - if ( rc ) - hvm_free_ioreq_mfn(s, false); - - return rc; -} - -static void hvm_ioreq_server_free_pages(struct hvm_ioreq_server *s) -{ - hvm_free_ioreq_mfn(s, true); - hvm_free_ioreq_mfn(s, false); -} - -static void hvm_ioreq_server_free_rangesets(struct hvm_ioreq_server *s) -{ - unsigned int i; - - for ( i = 0; i < NR_IO_RANGE_TYPES; i++ ) - rangeset_destroy(s->range[i]); -} - -static int hvm_ioreq_server_alloc_rangesets(struct hvm_ioreq_server *s, - ioservid_t id) -{ - unsigned int i; - int rc; - - for ( i = 0; i < NR_IO_RANGE_TYPES; i++ ) - { - char *name; - - rc = asprintf(&name, "ioreq_server %d %s", id, - (i == XEN_DMOP_IO_RANGE_PORT) ? "port" : - (i == XEN_DMOP_IO_RANGE_MEMORY) ? "memory" : - (i == XEN_DMOP_IO_RANGE_PCI) ? "pci" : - ""); - if ( rc ) - goto fail; - - s->range[i] = rangeset_new(s->target, name, - RANGESETF_prettyprint_hex); - - xfree(name); - - rc = -ENOMEM; - if ( !s->range[i] ) - goto fail; - - rangeset_limit(s->range[i], MAX_NR_IO_RANGES); - } - - return 0; - - fail: - hvm_ioreq_server_free_rangesets(s); - - return rc; -} - void arch_ioreq_server_enable(struct hvm_ioreq_server *s) { hvm_remove_ioreq_gfn(s, false); hvm_remove_ioreq_gfn(s, true); } -static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) -{ - struct hvm_ioreq_vcpu *sv; - - spin_lock(&s->lock); - - if ( s->enabled ) - goto done; - - arch_ioreq_server_enable(s); - - s->enabled = true; - - list_for_each_entry ( sv, - &s->ioreq_vcpu_list, - list_entry ) - hvm_update_ioreq_evtchn(s, sv); - - done: - spin_unlock(&s->lock); -} - void arch_ioreq_server_disable(struct hvm_ioreq_server *s) { hvm_add_ioreq_gfn(s, true); hvm_add_ioreq_gfn(s, false); } -static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) +/* Called when target domain is paused */ +void arch_ioreq_server_destroy(struct hvm_ioreq_server *s) { - spin_lock(&s->lock); - - if ( !s->enabled ) - goto done; - - arch_ioreq_server_disable(s); + p2m_set_ioreq_server(s->target, 0, s); +} - s->enabled = false; +/* Called with ioreq_server lock held */ +int arch_ioreq_server_map_mem_type(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags) +{ + return p2m_set_ioreq_server(d, flags, s); +} - done: - spin_unlock(&s->lock); +void arch_ioreq_server_map_mem_type_completed(struct domain *d, + struct hvm_ioreq_server *s, + uint32_t flags) +{ + if ( flags == 0 && read_atomic(&p2m_get_hostp2m(d)->ioreq.entry_count) ) + p2m_change_entry_type_global(d, p2m_ioreq_server, p2m_ram_rw); } -static int hvm_ioreq_server_init(struct hvm_ioreq_server *s, - struct domain *d, int bufioreq_handling, - ioservid_t id) +bool arch_ioreq_server_destroy_all(struct domain *d) { - struct domain *currd = current->domain; - struct vcpu *v; - int rc; + return relocate_portio_handler(d, 0xcf8, 0xcf8, 4); +} - s->target = d; +bool arch_ioreq_server_get_type_addr(const struct domain *d, + const ioreq_t *p, + uint8_t *type, + uint64_t *addr) +{ + unsigned int cf8 = d->arch.hvm.pci_cf8; - get_knownalive_domain(currd); - s->emulator = currd; + if ( p->type != IOREQ_TYPE_COPY && p->type != IOREQ_TYPE_PIO ) + return false; - spin_lock_init(&s->lock); - INIT_LIST_HEAD(&s->ioreq_vcpu_list); - spin_lock_init(&s->bufioreq_lock); + if ( p->type == IOREQ_TYPE_PIO && + (p->addr & ~3) == 0xcfc && + CF8_ENABLED(cf8) ) + { + unsigned int x86_fam, reg; + pci_sbdf_t sbdf; - s->ioreq.gfn = INVALID_GFN; - s->bufioreq.gfn = INVALID_GFN; - - rc = hvm_ioreq_server_alloc_rangesets(s, id); - if ( rc ) - return rc; - - s->bufioreq_handling = bufioreq_handling; - - for_each_vcpu ( d, v ) - { - rc = hvm_ioreq_server_add_vcpu(s, v); - if ( rc ) - goto fail_add; - } - - return 0; - - fail_add: - hvm_ioreq_server_remove_all_vcpus(s); - arch_ioreq_server_unmap_pages(s); - - hvm_ioreq_server_free_rangesets(s); - - put_domain(s->emulator); - return rc; -} - -static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) -{ - ASSERT(!s->enabled); - hvm_ioreq_server_remove_all_vcpus(s); - - /* - * NOTE: It is safe to call both arch_ioreq_server_unmap_pages() and - * hvm_ioreq_server_free_pages() in that order. - * This is because the former will do nothing if the pages - * are not mapped, leaving the page to be freed by the latter. - * However if the pages are mapped then the former will set - * the page_info pointer to NULL, meaning the latter will do - * nothing. - */ - arch_ioreq_server_unmap_pages(s); - hvm_ioreq_server_free_pages(s); - - hvm_ioreq_server_free_rangesets(s); - - put_domain(s->emulator); -} - -int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, - ioservid_t *id) -{ - struct hvm_ioreq_server *s; - unsigned int i; - int rc; - - if ( bufioreq_handling > HVM_IOREQSRV_BUFIOREQ_ATOMIC ) - return -EINVAL; - - s = xzalloc(struct hvm_ioreq_server); - if ( !s ) - return -ENOMEM; - - domain_pause(d); - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - for ( i = 0; i < MAX_NR_IOREQ_SERVERS; i++ ) - { - if ( !GET_IOREQ_SERVER(d, i) ) - break; - } - - rc = -ENOSPC; - if ( i >= MAX_NR_IOREQ_SERVERS ) - goto fail; - - /* - * It is safe to call set_ioreq_server() prior to - * hvm_ioreq_server_init() since the target domain is paused. - */ - set_ioreq_server(d, i, s); - - rc = hvm_ioreq_server_init(s, d, bufioreq_handling, i); - if ( rc ) - { - set_ioreq_server(d, i, NULL); - goto fail; - } - - if ( id ) - *id = i; - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - domain_unpause(d); - - return 0; - - fail: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - domain_unpause(d); - - xfree(s); - return rc; -} - -/* Called when target domain is paused */ -void arch_ioreq_server_destroy(struct hvm_ioreq_server *s) -{ - p2m_set_ioreq_server(s->target, 0, s); -} - -int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) -{ - struct hvm_ioreq_server *s; - int rc; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - domain_pause(d); - - arch_ioreq_server_destroy(s); - - hvm_ioreq_server_disable(s); - - /* - * It is safe to call hvm_ioreq_server_deinit() prior to - * set_ioreq_server() since the target domain is paused. - */ - hvm_ioreq_server_deinit(s); - set_ioreq_server(d, id, NULL); - - domain_unpause(d); - - xfree(s); - - rc = 0; - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, - unsigned long *ioreq_gfn, - unsigned long *bufioreq_gfn, - evtchn_port_t *bufioreq_port) -{ - struct hvm_ioreq_server *s; - int rc; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - if ( ioreq_gfn || bufioreq_gfn ) - { - rc = arch_ioreq_server_map_pages(s); - if ( rc ) - goto out; - } - - if ( ioreq_gfn ) - *ioreq_gfn = gfn_x(s->ioreq.gfn); - - if ( HANDLE_BUFIOREQ(s) ) - { - if ( bufioreq_gfn ) - *bufioreq_gfn = gfn_x(s->bufioreq.gfn); - - if ( bufioreq_port ) - *bufioreq_port = s->bufioreq_evtchn; - } - - rc = 0; - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, - unsigned long idx, mfn_t *mfn) -{ - struct hvm_ioreq_server *s; - int rc; - - ASSERT(is_hvm_domain(d)); - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - rc = hvm_ioreq_server_alloc_pages(s); - if ( rc ) - goto out; - - switch ( idx ) - { - case XENMEM_resource_ioreq_server_frame_bufioreq: - rc = -ENOENT; - if ( !HANDLE_BUFIOREQ(s) ) - goto out; - - *mfn = page_to_mfn(s->bufioreq.page); - rc = 0; - break; - - case XENMEM_resource_ioreq_server_frame_ioreq(0): - *mfn = page_to_mfn(s->ioreq.page); - rc = 0; - break; - - default: - rc = -EINVAL; - break; - } - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end) -{ - struct hvm_ioreq_server *s; - struct rangeset *r; - int rc; - - if ( start > end ) - return -EINVAL; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - switch ( type ) - { - case XEN_DMOP_IO_RANGE_PORT: - case XEN_DMOP_IO_RANGE_MEMORY: - case XEN_DMOP_IO_RANGE_PCI: - r = s->range[type]; - break; - - default: - r = NULL; - break; - } - - rc = -EINVAL; - if ( !r ) - goto out; - - rc = -EEXIST; - if ( rangeset_overlaps_range(r, start, end) ) - goto out; - - rc = rangeset_add_range(r, start, end); - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end) -{ - struct hvm_ioreq_server *s; - struct rangeset *r; - int rc; - - if ( start > end ) - return -EINVAL; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - switch ( type ) - { - case XEN_DMOP_IO_RANGE_PORT: - case XEN_DMOP_IO_RANGE_MEMORY: - case XEN_DMOP_IO_RANGE_PCI: - r = s->range[type]; - break; - - default: - r = NULL; - break; - } - - rc = -EINVAL; - if ( !r ) - goto out; - - rc = -ENOENT; - if ( !rangeset_contains_range(r, start, end) ) - goto out; - - rc = rangeset_remove_range(r, start, end); - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -/* Called with ioreq_server lock held */ -int arch_ioreq_server_map_mem_type(struct domain *d, - struct hvm_ioreq_server *s, - uint32_t flags) -{ - return p2m_set_ioreq_server(d, flags, s); -} - -void arch_ioreq_server_map_mem_type_completed(struct domain *d, - struct hvm_ioreq_server *s, - uint32_t flags) -{ - if ( flags == 0 && read_atomic(&p2m_get_hostp2m(d)->ioreq.entry_count) ) - p2m_change_entry_type_global(d, p2m_ioreq_server, p2m_ram_rw); -} - -/* - * Map or unmap an ioreq server to specific memory type. For now, only - * HVMMEM_ioreq_server is supported, and in the future new types can be - * introduced, e.g. HVMMEM_ioreq_serverX mapped to ioreq server X. And - * currently, only write operations are to be forwarded to an ioreq server. - * Support for the emulation of read operations can be added when an ioreq - * server has such requirement in the future. - */ -int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint32_t flags) -{ - struct hvm_ioreq_server *s; - int rc; - - if ( type != HVMMEM_ioreq_server ) - return -EINVAL; - - if ( flags & ~XEN_DMOP_IOREQ_MEM_ACCESS_WRITE ) - return -EINVAL; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - rc = arch_ioreq_server_map_mem_type(d, s, flags); - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - if ( rc == 0 ) - arch_ioreq_server_map_mem_type_completed(d, s, flags); - - return rc; -} - -int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, - bool enabled) -{ - struct hvm_ioreq_server *s; - int rc; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - s = get_ioreq_server(d, id); - - rc = -ENOENT; - if ( !s ) - goto out; - - rc = -EPERM; - if ( s->emulator != current->domain ) - goto out; - - domain_pause(d); - - if ( enabled ) - hvm_ioreq_server_enable(s); - else - hvm_ioreq_server_disable(s); - - domain_unpause(d); - - rc = 0; - - out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - return rc; -} - -int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) -{ - struct hvm_ioreq_server *s; - unsigned int id; - int rc; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - rc = hvm_ioreq_server_add_vcpu(s, v); - if ( rc ) - goto fail; - } - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return 0; - - fail: - while ( ++id != MAX_NR_IOREQ_SERVERS ) - { - s = GET_IOREQ_SERVER(d, id); - - if ( !s ) - continue; - - hvm_ioreq_server_remove_vcpu(s, v); - } - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); - - return rc; -} - -void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) -{ - struct hvm_ioreq_server *s; - unsigned int id; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - FOR_EACH_IOREQ_SERVER(d, id, s) - hvm_ioreq_server_remove_vcpu(s, v); - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); -} - -bool arch_ioreq_server_destroy_all(struct domain *d) -{ - return relocate_portio_handler(d, 0xcf8, 0xcf8, 4); -} - -void hvm_destroy_all_ioreq_servers(struct domain *d) -{ - struct hvm_ioreq_server *s; - unsigned int id; - - if ( !arch_ioreq_server_destroy_all(d) ) - return; - - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); - - /* No need to domain_pause() as the domain is being torn down */ - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - hvm_ioreq_server_disable(s); - - /* - * It is safe to call hvm_ioreq_server_deinit() prior to - * set_ioreq_server() since the target domain is being destroyed. - */ - hvm_ioreq_server_deinit(s); - set_ioreq_server(d, id, NULL); - - xfree(s); - } - - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); -} - -bool arch_ioreq_server_get_type_addr(const struct domain *d, - const ioreq_t *p, - uint8_t *type, - uint64_t *addr) -{ - unsigned int cf8 = d->arch.hvm.pci_cf8; - - if ( p->type != IOREQ_TYPE_COPY && p->type != IOREQ_TYPE_PIO ) - return false; - - if ( p->type == IOREQ_TYPE_PIO && - (p->addr & ~3) == 0xcfc && - CF8_ENABLED(cf8) ) - { - unsigned int x86_fam, reg; - pci_sbdf_t sbdf; - - reg = hvm_pci_decode_addr(cf8, p->addr, &sbdf); + reg = hvm_pci_decode_addr(cf8, p->addr, &sbdf); /* PCI config data cycle */ *type = XEN_DMOP_IO_RANGE_PCI; @@ -1331,233 +319,6 @@ bool arch_ioreq_server_get_type_addr(const struct domain *d, return true; } -struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p) -{ - struct hvm_ioreq_server *s; - uint8_t type; - uint64_t addr; - unsigned int id; - - if ( !arch_ioreq_server_get_type_addr(d, p, &type, &addr) ) - return NULL; - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - struct rangeset *r; - - if ( !s->enabled ) - continue; - - r = s->range[type]; - - switch ( type ) - { - unsigned long start, end; - - case XEN_DMOP_IO_RANGE_PORT: - start = addr; - end = start + p->size - 1; - if ( rangeset_contains_range(r, start, end) ) - return s; - - break; - - case XEN_DMOP_IO_RANGE_MEMORY: - start = hvm_mmio_first_byte(p); - end = hvm_mmio_last_byte(p); - - if ( rangeset_contains_range(r, start, end) ) - return s; - - break; - - case XEN_DMOP_IO_RANGE_PCI: - if ( rangeset_contains_singleton(r, addr >> 32) ) - { - p->type = IOREQ_TYPE_PCI_CONFIG; - p->addr = addr; - return s; - } - - break; - } - } - - return NULL; -} - -static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) -{ - struct domain *d = current->domain; - struct hvm_ioreq_page *iorp; - buffered_iopage_t *pg; - buf_ioreq_t bp = { .data = p->data, - .addr = p->addr, - .type = p->type, - .dir = p->dir }; - /* Timeoffset sends 64b data, but no address. Use two consecutive slots. */ - int qw = 0; - - /* Ensure buffered_iopage fits in a page */ - BUILD_BUG_ON(sizeof(buffered_iopage_t) > PAGE_SIZE); - - iorp = &s->bufioreq; - pg = iorp->va; - - if ( !pg ) - return IOREQ_STATUS_UNHANDLED; - - /* - * Return 0 for the cases we can't deal with: - * - 'addr' is only a 20-bit field, so we cannot address beyond 1MB - * - we cannot buffer accesses to guest memory buffers, as the guest - * may expect the memory buffer to be synchronously accessed - * - the count field is usually used with data_is_ptr and since we don't - * support data_is_ptr we do not waste space for the count field either - */ - if ( (p->addr > 0xffffful) || p->data_is_ptr || (p->count != 1) ) - return 0; - - switch ( p->size ) - { - case 1: - bp.size = 0; - break; - case 2: - bp.size = 1; - break; - case 4: - bp.size = 2; - break; - case 8: - bp.size = 3; - qw = 1; - break; - default: - gdprintk(XENLOG_WARNING, "unexpected ioreq size: %u\n", p->size); - return IOREQ_STATUS_UNHANDLED; - } - - spin_lock(&s->bufioreq_lock); - - if ( (pg->ptrs.write_pointer - pg->ptrs.read_pointer) >= - (IOREQ_BUFFER_SLOT_NUM - qw) ) - { - /* The queue is full: send the iopacket through the normal path. */ - spin_unlock(&s->bufioreq_lock); - return IOREQ_STATUS_UNHANDLED; - } - - pg->buf_ioreq[pg->ptrs.write_pointer % IOREQ_BUFFER_SLOT_NUM] = bp; - - if ( qw ) - { - bp.data = p->data >> 32; - pg->buf_ioreq[(pg->ptrs.write_pointer+1) % IOREQ_BUFFER_SLOT_NUM] = bp; - } - - /* Make the ioreq_t visible /before/ write_pointer. */ - smp_wmb(); - pg->ptrs.write_pointer += qw ? 2 : 1; - - /* Canonicalize read/write pointers to prevent their overflow. */ - while ( (s->bufioreq_handling == HVM_IOREQSRV_BUFIOREQ_ATOMIC) && - qw++ < IOREQ_BUFFER_SLOT_NUM && - pg->ptrs.read_pointer >= IOREQ_BUFFER_SLOT_NUM ) - { - union bufioreq_pointers old = pg->ptrs, new; - unsigned int n = old.read_pointer / IOREQ_BUFFER_SLOT_NUM; - - new.read_pointer = old.read_pointer - n * IOREQ_BUFFER_SLOT_NUM; - new.write_pointer = old.write_pointer - n * IOREQ_BUFFER_SLOT_NUM; - cmpxchg(&pg->ptrs.full, old.full, new.full); - } - - notify_via_xen_event_channel(d, s->bufioreq_evtchn); - spin_unlock(&s->bufioreq_lock); - - return IOREQ_STATUS_HANDLED; -} - -int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, - bool buffered) -{ - struct vcpu *curr = current; - struct domain *d = curr->domain; - struct hvm_ioreq_vcpu *sv; - - ASSERT(s); - - if ( buffered ) - return hvm_send_buffered_ioreq(s, proto_p); - - if ( unlikely(!vcpu_start_shutdown_deferral(curr)) ) - return IOREQ_STATUS_RETRY; - - list_for_each_entry ( sv, - &s->ioreq_vcpu_list, - list_entry ) - { - if ( sv->vcpu == curr ) - { - evtchn_port_t port = sv->ioreq_evtchn; - ioreq_t *p = get_ioreq(s, curr); - - if ( unlikely(p->state != STATE_IOREQ_NONE) ) - { - gprintk(XENLOG_ERR, "device model set bad IO state %d\n", - p->state); - break; - } - - if ( unlikely(p->vp_eport != port) ) - { - gprintk(XENLOG_ERR, "device model set bad event channel %d\n", - p->vp_eport); - break; - } - - proto_p->state = STATE_IOREQ_NONE; - proto_p->vp_eport = port; - *p = *proto_p; - - prepare_wait_on_xen_event_channel(port); - - /* - * Following happens /after/ blocking and setting up ioreq - * contents. prepare_wait_on_xen_event_channel() is an implicit - * barrier. - */ - p->state = STATE_IOREQ_READY; - notify_via_xen_event_channel(d, port); - - sv->pending = true; - return IOREQ_STATUS_RETRY; - } - } - - return IOREQ_STATUS_UNHANDLED; -} - -unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) -{ - struct domain *d = current->domain; - struct hvm_ioreq_server *s; - unsigned int id, failed = 0; - - FOR_EACH_IOREQ_SERVER(d, id, s) - { - if ( !s->enabled ) - continue; - - if ( hvm_send_ioreq(s, p, buffered) == IOREQ_STATUS_UNHANDLED ) - failed++; - } - - return failed; -} - static int hvm_access_cf8( int dir, unsigned int port, unsigned int bytes, uint32_t *val) { @@ -1575,13 +336,6 @@ void arch_ioreq_domain_init(struct domain *d) register_portio_handler(d, 0xcf8, 4, hvm_access_cf8); } -void hvm_ioreq_init(struct domain *d) -{ - spin_lock_init(&d->arch.hvm.ioreq_server.lock); - - arch_ioreq_domain_init(d); -} - /* * Local variables: * mode: C diff --git a/xen/arch/x86/hvm/stdvga.c b/xen/arch/x86/hvm/stdvga.c index e267513..fd7cadb 100644 --- a/xen/arch/x86/hvm/stdvga.c +++ b/xen/arch/x86/hvm/stdvga.c @@ -27,10 +27,10 @@ * can have side effects. */ +#include #include #include #include -#include #include #include #include diff --git a/xen/arch/x86/hvm/vmx/vvmx.c b/xen/arch/x86/hvm/vmx/vvmx.c index 3a37e9e..0ddb6a4 100644 --- a/xen/arch/x86/hvm/vmx/vvmx.c +++ b/xen/arch/x86/hvm/vmx/vvmx.c @@ -19,10 +19,11 @@ * */ +#include + #include #include #include -#include #include #include #include diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c index 63e9fae..59eb5c8 100644 --- a/xen/arch/x86/mm.c +++ b/xen/arch/x86/mm.c @@ -100,6 +100,7 @@ */ #include +#include #include #include #include @@ -140,7 +141,6 @@ #include #include #include -#include #include #include diff --git a/xen/arch/x86/mm/shadow/common.c b/xen/arch/x86/mm/shadow/common.c index df95c4d..4a88824 100644 --- a/xen/arch/x86/mm/shadow/common.c +++ b/xen/arch/x86/mm/shadow/common.c @@ -20,6 +20,7 @@ * along with this program; If not, see . */ +#include #include #include #include @@ -34,7 +35,6 @@ #include #include #include -#include #include #include "private.h" diff --git a/xen/common/Kconfig b/xen/common/Kconfig index b5c91a1..cf32a07 100644 --- a/xen/common/Kconfig +++ b/xen/common/Kconfig @@ -139,6 +139,9 @@ config HYPFS_CONFIG Disable this option in case you want to spare some memory or you want to hide the .config contents from dom0. +config IOREQ_SERVER + bool + config KEXEC bool "kexec support" default y diff --git a/xen/common/Makefile b/xen/common/Makefile index d751315..2e3c234 100644 --- a/xen/common/Makefile +++ b/xen/common/Makefile @@ -14,6 +14,7 @@ obj-$(CONFIG_GRANT_TABLE) += grant_table.o obj-y += guestcopy.o obj-bin-y += gunzip.init.o obj-$(CONFIG_HYPFS) += hypfs.o +obj-$(CONFIG_IOREQ_SERVER) += ioreq.o obj-y += irq.o obj-y += kernel.o obj-y += keyhandler.o diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c new file mode 100644 index 0000000..4e7d91b --- /dev/null +++ b/xen/common/ioreq.c @@ -0,0 +1,1290 @@ +/* + * ioreq.c: hardware virtual machine I/O emulation + * + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include +#include + +static void set_ioreq_server(struct domain *d, unsigned int id, + struct hvm_ioreq_server *s) +{ + ASSERT(id < MAX_NR_IOREQ_SERVERS); + ASSERT(!s || !d->arch.hvm.ioreq_server.server[id]); + + d->arch.hvm.ioreq_server.server[id] = s; +} + +#define GET_IOREQ_SERVER(d, id) \ + (d)->arch.hvm.ioreq_server.server[id] + +static struct hvm_ioreq_server *get_ioreq_server(const struct domain *d, + unsigned int id) +{ + if ( id >= MAX_NR_IOREQ_SERVERS ) + return NULL; + + return GET_IOREQ_SERVER(d, id); +} + +/* + * Iterate over all possible ioreq servers. + * + * NOTE: The iteration is backwards such that more recently created + * ioreq servers are favoured in hvm_select_ioreq_server(). + * This is a semantic that previously existed when ioreq servers + * were held in a linked list. + */ +#define FOR_EACH_IOREQ_SERVER(d, id, s) \ + for ( (id) = MAX_NR_IOREQ_SERVERS; (id) != 0; ) \ + if ( !(s = GET_IOREQ_SERVER(d, --(id))) ) \ + continue; \ + else + +static ioreq_t *get_ioreq(struct hvm_ioreq_server *s, struct vcpu *v) +{ + shared_iopage_t *p = s->ioreq.va; + + ASSERT((v == current) || !vcpu_runnable(v)); + ASSERT(p != NULL); + + return &p->vcpu_ioreq[v->vcpu_id]; +} + +static struct hvm_ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, + struct hvm_ioreq_server **srvp) +{ + struct domain *d = v->domain; + struct hvm_ioreq_server *s; + unsigned int id; + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + struct hvm_ioreq_vcpu *sv; + + list_for_each_entry ( sv, + &s->ioreq_vcpu_list, + list_entry ) + { + if ( sv->vcpu == v && sv->pending ) + { + if ( srvp ) + *srvp = s; + return sv; + } + } + } + + return NULL; +} + +bool hvm_io_pending(struct vcpu *v) +{ + return get_pending_vcpu(v, NULL); +} + +static bool hvm_wait_for_io(struct hvm_ioreq_vcpu *sv, ioreq_t *p) +{ + unsigned int prev_state = STATE_IOREQ_NONE; + unsigned int state = p->state; + uint64_t data = ~0; + + smp_rmb(); + + /* + * The only reason we should see this condition be false is when an + * emulator dying races with I/O being requested. + */ + while ( likely(state != STATE_IOREQ_NONE) ) + { + if ( unlikely(state < prev_state) ) + { + gdprintk(XENLOG_ERR, "Weird HVM ioreq state transition %u -> %u\n", + prev_state, state); + sv->pending = false; + domain_crash(sv->vcpu->domain); + return false; /* bail */ + } + + switch ( prev_state = state ) + { + case STATE_IORESP_READY: /* IORESP_READY -> NONE */ + p->state = STATE_IOREQ_NONE; + data = p->data; + break; + + case STATE_IOREQ_READY: /* IOREQ_{READY,INPROCESS} -> IORESP_READY */ + case STATE_IOREQ_INPROCESS: + wait_on_xen_event_channel(sv->ioreq_evtchn, + ({ state = p->state; + smp_rmb(); + state != prev_state; })); + continue; + + default: + gdprintk(XENLOG_ERR, "Weird HVM iorequest state %u\n", state); + sv->pending = false; + domain_crash(sv->vcpu->domain); + return false; /* bail */ + } + + break; + } + + p = &sv->vcpu->arch.hvm.hvm_io.io_req; + if ( hvm_ioreq_needs_completion(p) ) + p->data = data; + + sv->pending = false; + + return true; +} + +bool handle_hvm_io_completion(struct vcpu *v) +{ + struct domain *d = v->domain; + struct hvm_vcpu_io *vio = &v->arch.hvm.hvm_io; + struct hvm_ioreq_server *s; + struct hvm_ioreq_vcpu *sv; + enum hvm_io_completion io_completion; + + if ( has_vpci(d) && vpci_process_pending(v) ) + { + raise_softirq(SCHEDULE_SOFTIRQ); + return false; + } + + sv = get_pending_vcpu(v, &s); + if ( sv && !hvm_wait_for_io(sv, get_ioreq(s, v)) ) + return false; + + vio->io_req.state = hvm_ioreq_needs_completion(&vio->io_req) ? + STATE_IORESP_READY : STATE_IOREQ_NONE; + + msix_write_completion(v); + vcpu_end_shutdown_deferral(v); + + io_completion = vio->io_completion; + vio->io_completion = HVMIO_no_completion; + + switch ( io_completion ) + { + case HVMIO_no_completion: + break; + + case HVMIO_mmio_completion: + return arch_ioreq_complete_mmio(); + + case HVMIO_pio_completion: + return handle_pio(vio->io_req.addr, vio->io_req.size, + vio->io_req.dir); + + default: + return arch_vcpu_ioreq_completion(io_completion); + } + + return true; +} + +static int hvm_alloc_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) +{ + struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct page_info *page; + + if ( iorp->page ) + { + /* + * If a guest frame has already been mapped (which may happen + * on demand if hvm_get_ioreq_server_info() is called), then + * allocating a page is not permitted. + */ + if ( !gfn_eq(iorp->gfn, INVALID_GFN) ) + return -EPERM; + + return 0; + } + + page = alloc_domheap_page(s->target, MEMF_no_refcount); + + if ( !page ) + return -ENOMEM; + + if ( !get_page_and_type(page, s->target, PGT_writable_page) ) + { + /* + * The domain can't possibly know about this page yet, so failure + * here is a clear indication of something fishy going on. + */ + domain_crash(s->emulator); + return -ENODATA; + } + + iorp->va = __map_domain_page_global(page); + if ( !iorp->va ) + goto fail; + + iorp->page = page; + clear_page(iorp->va); + return 0; + + fail: + put_page_alloc_ref(page); + put_page_and_type(page); + + return -ENOMEM; +} + +static void hvm_free_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) +{ + struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct page_info *page = iorp->page; + + if ( !page ) + return; + + iorp->page = NULL; + + unmap_domain_page_global(iorp->va); + iorp->va = NULL; + + put_page_alloc_ref(page); + put_page_and_type(page); +} + +bool is_ioreq_server_page(struct domain *d, const struct page_info *page) +{ + const struct hvm_ioreq_server *s; + unsigned int id; + bool found = false; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + if ( (s->ioreq.page == page) || (s->bufioreq.page == page) ) + { + found = true; + break; + } + } + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return found; +} + +static void hvm_update_ioreq_evtchn(struct hvm_ioreq_server *s, + struct hvm_ioreq_vcpu *sv) +{ + ASSERT(spin_is_locked(&s->lock)); + + if ( s->ioreq.va != NULL ) + { + ioreq_t *p = get_ioreq(s, sv->vcpu); + + p->vp_eport = sv->ioreq_evtchn; + } +} + +static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, + struct vcpu *v) +{ + struct hvm_ioreq_vcpu *sv; + int rc; + + sv = xzalloc(struct hvm_ioreq_vcpu); + + rc = -ENOMEM; + if ( !sv ) + goto fail1; + + spin_lock(&s->lock); + + rc = alloc_unbound_xen_event_channel(v->domain, v->vcpu_id, + s->emulator->domain_id, NULL); + if ( rc < 0 ) + goto fail2; + + sv->ioreq_evtchn = rc; + + if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) + { + rc = alloc_unbound_xen_event_channel(v->domain, 0, + s->emulator->domain_id, NULL); + if ( rc < 0 ) + goto fail3; + + s->bufioreq_evtchn = rc; + } + + sv->vcpu = v; + + list_add(&sv->list_entry, &s->ioreq_vcpu_list); + + if ( s->enabled ) + hvm_update_ioreq_evtchn(s, sv); + + spin_unlock(&s->lock); + return 0; + + fail3: + free_xen_event_channel(v->domain, sv->ioreq_evtchn); + + fail2: + spin_unlock(&s->lock); + xfree(sv); + + fail1: + return rc; +} + +static void hvm_ioreq_server_remove_vcpu(struct hvm_ioreq_server *s, + struct vcpu *v) +{ + struct hvm_ioreq_vcpu *sv; + + spin_lock(&s->lock); + + list_for_each_entry ( sv, + &s->ioreq_vcpu_list, + list_entry ) + { + if ( sv->vcpu != v ) + continue; + + list_del(&sv->list_entry); + + if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) + free_xen_event_channel(v->domain, s->bufioreq_evtchn); + + free_xen_event_channel(v->domain, sv->ioreq_evtchn); + + xfree(sv); + break; + } + + spin_unlock(&s->lock); +} + +static void hvm_ioreq_server_remove_all_vcpus(struct hvm_ioreq_server *s) +{ + struct hvm_ioreq_vcpu *sv, *next; + + spin_lock(&s->lock); + + list_for_each_entry_safe ( sv, + next, + &s->ioreq_vcpu_list, + list_entry ) + { + struct vcpu *v = sv->vcpu; + + list_del(&sv->list_entry); + + if ( v->vcpu_id == 0 && HANDLE_BUFIOREQ(s) ) + free_xen_event_channel(v->domain, s->bufioreq_evtchn); + + free_xen_event_channel(v->domain, sv->ioreq_evtchn); + + xfree(sv); + } + + spin_unlock(&s->lock); +} + +static int hvm_ioreq_server_alloc_pages(struct hvm_ioreq_server *s) +{ + int rc; + + rc = hvm_alloc_ioreq_mfn(s, false); + + if ( !rc && (s->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) ) + rc = hvm_alloc_ioreq_mfn(s, true); + + if ( rc ) + hvm_free_ioreq_mfn(s, false); + + return rc; +} + +static void hvm_ioreq_server_free_pages(struct hvm_ioreq_server *s) +{ + hvm_free_ioreq_mfn(s, true); + hvm_free_ioreq_mfn(s, false); +} + +static void hvm_ioreq_server_free_rangesets(struct hvm_ioreq_server *s) +{ + unsigned int i; + + for ( i = 0; i < NR_IO_RANGE_TYPES; i++ ) + rangeset_destroy(s->range[i]); +} + +static int hvm_ioreq_server_alloc_rangesets(struct hvm_ioreq_server *s, + ioservid_t id) +{ + unsigned int i; + int rc; + + for ( i = 0; i < NR_IO_RANGE_TYPES; i++ ) + { + char *name; + + rc = asprintf(&name, "ioreq_server %d %s", id, + (i == XEN_DMOP_IO_RANGE_PORT) ? "port" : + (i == XEN_DMOP_IO_RANGE_MEMORY) ? "memory" : + (i == XEN_DMOP_IO_RANGE_PCI) ? "pci" : + ""); + if ( rc ) + goto fail; + + s->range[i] = rangeset_new(s->target, name, + RANGESETF_prettyprint_hex); + + xfree(name); + + rc = -ENOMEM; + if ( !s->range[i] ) + goto fail; + + rangeset_limit(s->range[i], MAX_NR_IO_RANGES); + } + + return 0; + + fail: + hvm_ioreq_server_free_rangesets(s); + + return rc; +} + +static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) +{ + struct hvm_ioreq_vcpu *sv; + + spin_lock(&s->lock); + + if ( s->enabled ) + goto done; + + arch_ioreq_server_enable(s); + + s->enabled = true; + + list_for_each_entry ( sv, + &s->ioreq_vcpu_list, + list_entry ) + hvm_update_ioreq_evtchn(s, sv); + + done: + spin_unlock(&s->lock); +} + +static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) +{ + spin_lock(&s->lock); + + if ( !s->enabled ) + goto done; + + arch_ioreq_server_disable(s); + + s->enabled = false; + + done: + spin_unlock(&s->lock); +} + +static int hvm_ioreq_server_init(struct hvm_ioreq_server *s, + struct domain *d, int bufioreq_handling, + ioservid_t id) +{ + struct domain *currd = current->domain; + struct vcpu *v; + int rc; + + s->target = d; + + get_knownalive_domain(currd); + s->emulator = currd; + + spin_lock_init(&s->lock); + INIT_LIST_HEAD(&s->ioreq_vcpu_list); + spin_lock_init(&s->bufioreq_lock); + + s->ioreq.gfn = INVALID_GFN; + s->bufioreq.gfn = INVALID_GFN; + + rc = hvm_ioreq_server_alloc_rangesets(s, id); + if ( rc ) + return rc; + + s->bufioreq_handling = bufioreq_handling; + + for_each_vcpu ( d, v ) + { + rc = hvm_ioreq_server_add_vcpu(s, v); + if ( rc ) + goto fail_add; + } + + return 0; + + fail_add: + hvm_ioreq_server_remove_all_vcpus(s); + arch_ioreq_server_unmap_pages(s); + + hvm_ioreq_server_free_rangesets(s); + + put_domain(s->emulator); + return rc; +} + +static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) +{ + ASSERT(!s->enabled); + hvm_ioreq_server_remove_all_vcpus(s); + + /* + * NOTE: It is safe to call both arch_ioreq_server_unmap_pages() and + * hvm_ioreq_server_free_pages() in that order. + * This is because the former will do nothing if the pages + * are not mapped, leaving the page to be freed by the latter. + * However if the pages are mapped then the former will set + * the page_info pointer to NULL, meaning the latter will do + * nothing. + */ + arch_ioreq_server_unmap_pages(s); + hvm_ioreq_server_free_pages(s); + + hvm_ioreq_server_free_rangesets(s); + + put_domain(s->emulator); +} + +int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, + ioservid_t *id) +{ + struct hvm_ioreq_server *s; + unsigned int i; + int rc; + + if ( bufioreq_handling > HVM_IOREQSRV_BUFIOREQ_ATOMIC ) + return -EINVAL; + + s = xzalloc(struct hvm_ioreq_server); + if ( !s ) + return -ENOMEM; + + domain_pause(d); + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + for ( i = 0; i < MAX_NR_IOREQ_SERVERS; i++ ) + { + if ( !GET_IOREQ_SERVER(d, i) ) + break; + } + + rc = -ENOSPC; + if ( i >= MAX_NR_IOREQ_SERVERS ) + goto fail; + + /* + * It is safe to call set_ioreq_server() prior to + * hvm_ioreq_server_init() since the target domain is paused. + */ + set_ioreq_server(d, i, s); + + rc = hvm_ioreq_server_init(s, d, bufioreq_handling, i); + if ( rc ) + { + set_ioreq_server(d, i, NULL); + goto fail; + } + + if ( id ) + *id = i; + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + domain_unpause(d); + + return 0; + + fail: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + domain_unpause(d); + + xfree(s); + return rc; +} + +int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) +{ + struct hvm_ioreq_server *s; + int rc; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + domain_pause(d); + + arch_ioreq_server_destroy(s); + + hvm_ioreq_server_disable(s); + + /* + * It is safe to call hvm_ioreq_server_deinit() prior to + * set_ioreq_server() since the target domain is paused. + */ + hvm_ioreq_server_deinit(s); + set_ioreq_server(d, id, NULL); + + domain_unpause(d); + + xfree(s); + + rc = 0; + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, + unsigned long *ioreq_gfn, + unsigned long *bufioreq_gfn, + evtchn_port_t *bufioreq_port) +{ + struct hvm_ioreq_server *s; + int rc; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + if ( ioreq_gfn || bufioreq_gfn ) + { + rc = arch_ioreq_server_map_pages(s); + if ( rc ) + goto out; + } + + if ( ioreq_gfn ) + *ioreq_gfn = gfn_x(s->ioreq.gfn); + + if ( HANDLE_BUFIOREQ(s) ) + { + if ( bufioreq_gfn ) + *bufioreq_gfn = gfn_x(s->bufioreq.gfn); + + if ( bufioreq_port ) + *bufioreq_port = s->bufioreq_evtchn; + } + + rc = 0; + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, + unsigned long idx, mfn_t *mfn) +{ + struct hvm_ioreq_server *s; + int rc; + + ASSERT(is_hvm_domain(d)); + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + rc = hvm_ioreq_server_alloc_pages(s); + if ( rc ) + goto out; + + switch ( idx ) + { + case XENMEM_resource_ioreq_server_frame_bufioreq: + rc = -ENOENT; + if ( !HANDLE_BUFIOREQ(s) ) + goto out; + + *mfn = page_to_mfn(s->bufioreq.page); + rc = 0; + break; + + case XENMEM_resource_ioreq_server_frame_ioreq(0): + *mfn = page_to_mfn(s->ioreq.page); + rc = 0; + break; + + default: + rc = -EINVAL; + break; + } + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint64_t start, + uint64_t end) +{ + struct hvm_ioreq_server *s; + struct rangeset *r; + int rc; + + if ( start > end ) + return -EINVAL; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + switch ( type ) + { + case XEN_DMOP_IO_RANGE_PORT: + case XEN_DMOP_IO_RANGE_MEMORY: + case XEN_DMOP_IO_RANGE_PCI: + r = s->range[type]; + break; + + default: + r = NULL; + break; + } + + rc = -EINVAL; + if ( !r ) + goto out; + + rc = -EEXIST; + if ( rangeset_overlaps_range(r, start, end) ) + goto out; + + rc = rangeset_add_range(r, start, end); + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint64_t start, + uint64_t end) +{ + struct hvm_ioreq_server *s; + struct rangeset *r; + int rc; + + if ( start > end ) + return -EINVAL; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + switch ( type ) + { + case XEN_DMOP_IO_RANGE_PORT: + case XEN_DMOP_IO_RANGE_MEMORY: + case XEN_DMOP_IO_RANGE_PCI: + r = s->range[type]; + break; + + default: + r = NULL; + break; + } + + rc = -EINVAL; + if ( !r ) + goto out; + + rc = -ENOENT; + if ( !rangeset_contains_range(r, start, end) ) + goto out; + + rc = rangeset_remove_range(r, start, end); + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +/* + * Map or unmap an ioreq server to specific memory type. For now, only + * HVMMEM_ioreq_server is supported, and in the future new types can be + * introduced, e.g. HVMMEM_ioreq_serverX mapped to ioreq server X. And + * currently, only write operations are to be forwarded to an ioreq server. + * Support for the emulation of read operations can be added when an ioreq + * server has such requirement in the future. + */ +int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint32_t flags) +{ + struct hvm_ioreq_server *s; + int rc; + + if ( type != HVMMEM_ioreq_server ) + return -EINVAL; + + if ( flags & ~XEN_DMOP_IOREQ_MEM_ACCESS_WRITE ) + return -EINVAL; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + rc = arch_ioreq_server_map_mem_type(d, s, flags); + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + if ( rc == 0 ) + arch_ioreq_server_map_mem_type_completed(d, s, flags); + + return rc; +} + +int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, + bool enabled) +{ + struct hvm_ioreq_server *s; + int rc; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + s = get_ioreq_server(d, id); + + rc = -ENOENT; + if ( !s ) + goto out; + + rc = -EPERM; + if ( s->emulator != current->domain ) + goto out; + + domain_pause(d); + + if ( enabled ) + hvm_ioreq_server_enable(s); + else + hvm_ioreq_server_disable(s); + + domain_unpause(d); + + rc = 0; + + out: + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + return rc; +} + +int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) +{ + struct hvm_ioreq_server *s; + unsigned int id; + int rc; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + rc = hvm_ioreq_server_add_vcpu(s, v); + if ( rc ) + goto fail; + } + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return 0; + + fail: + while ( ++id != MAX_NR_IOREQ_SERVERS ) + { + s = GET_IOREQ_SERVER(d, id); + + if ( !s ) + continue; + + hvm_ioreq_server_remove_vcpu(s, v); + } + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + + return rc; +} + +void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) +{ + struct hvm_ioreq_server *s; + unsigned int id; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + FOR_EACH_IOREQ_SERVER(d, id, s) + hvm_ioreq_server_remove_vcpu(s, v); + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); +} + +void hvm_destroy_all_ioreq_servers(struct domain *d) +{ + struct hvm_ioreq_server *s; + unsigned int id; + + if ( !arch_ioreq_server_destroy_all(d) ) + return; + + spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + + /* No need to domain_pause() as the domain is being torn down */ + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + hvm_ioreq_server_disable(s); + + /* + * It is safe to call hvm_ioreq_server_deinit() prior to + * set_ioreq_server() since the target domain is being destroyed. + */ + hvm_ioreq_server_deinit(s); + set_ioreq_server(d, id, NULL); + + xfree(s); + } + + spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); +} + +struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, + ioreq_t *p) +{ + struct hvm_ioreq_server *s; + uint8_t type; + uint64_t addr; + unsigned int id; + + if ( !arch_ioreq_server_get_type_addr(d, p, &type, &addr) ) + return NULL; + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + struct rangeset *r; + + if ( !s->enabled ) + continue; + + r = s->range[type]; + + switch ( type ) + { + unsigned long start, end; + + case XEN_DMOP_IO_RANGE_PORT: + start = addr; + end = start + p->size - 1; + if ( rangeset_contains_range(r, start, end) ) + return s; + + break; + + case XEN_DMOP_IO_RANGE_MEMORY: + start = hvm_mmio_first_byte(p); + end = hvm_mmio_last_byte(p); + + if ( rangeset_contains_range(r, start, end) ) + return s; + + break; + + case XEN_DMOP_IO_RANGE_PCI: + if ( rangeset_contains_singleton(r, addr >> 32) ) + { + p->type = IOREQ_TYPE_PCI_CONFIG; + p->addr = addr; + return s; + } + + break; + } + } + + return NULL; +} + +static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) +{ + struct domain *d = current->domain; + struct hvm_ioreq_page *iorp; + buffered_iopage_t *pg; + buf_ioreq_t bp = { .data = p->data, + .addr = p->addr, + .type = p->type, + .dir = p->dir }; + /* Timeoffset sends 64b data, but no address. Use two consecutive slots. */ + int qw = 0; + + /* Ensure buffered_iopage fits in a page */ + BUILD_BUG_ON(sizeof(buffered_iopage_t) > PAGE_SIZE); + + iorp = &s->bufioreq; + pg = iorp->va; + + if ( !pg ) + return IOREQ_STATUS_UNHANDLED; + + /* + * Return 0 for the cases we can't deal with: + * - 'addr' is only a 20-bit field, so we cannot address beyond 1MB + * - we cannot buffer accesses to guest memory buffers, as the guest + * may expect the memory buffer to be synchronously accessed + * - the count field is usually used with data_is_ptr and since we don't + * support data_is_ptr we do not waste space for the count field either + */ + if ( (p->addr > 0xffffful) || p->data_is_ptr || (p->count != 1) ) + return 0; + + switch ( p->size ) + { + case 1: + bp.size = 0; + break; + case 2: + bp.size = 1; + break; + case 4: + bp.size = 2; + break; + case 8: + bp.size = 3; + qw = 1; + break; + default: + gdprintk(XENLOG_WARNING, "unexpected ioreq size: %u\n", p->size); + return IOREQ_STATUS_UNHANDLED; + } + + spin_lock(&s->bufioreq_lock); + + if ( (pg->ptrs.write_pointer - pg->ptrs.read_pointer) >= + (IOREQ_BUFFER_SLOT_NUM - qw) ) + { + /* The queue is full: send the iopacket through the normal path. */ + spin_unlock(&s->bufioreq_lock); + return IOREQ_STATUS_UNHANDLED; + } + + pg->buf_ioreq[pg->ptrs.write_pointer % IOREQ_BUFFER_SLOT_NUM] = bp; + + if ( qw ) + { + bp.data = p->data >> 32; + pg->buf_ioreq[(pg->ptrs.write_pointer+1) % IOREQ_BUFFER_SLOT_NUM] = bp; + } + + /* Make the ioreq_t visible /before/ write_pointer. */ + smp_wmb(); + pg->ptrs.write_pointer += qw ? 2 : 1; + + /* Canonicalize read/write pointers to prevent their overflow. */ + while ( (s->bufioreq_handling == HVM_IOREQSRV_BUFIOREQ_ATOMIC) && + qw++ < IOREQ_BUFFER_SLOT_NUM && + pg->ptrs.read_pointer >= IOREQ_BUFFER_SLOT_NUM ) + { + union bufioreq_pointers old = pg->ptrs, new; + unsigned int n = old.read_pointer / IOREQ_BUFFER_SLOT_NUM; + + new.read_pointer = old.read_pointer - n * IOREQ_BUFFER_SLOT_NUM; + new.write_pointer = old.write_pointer - n * IOREQ_BUFFER_SLOT_NUM; + cmpxchg(&pg->ptrs.full, old.full, new.full); + } + + notify_via_xen_event_channel(d, s->bufioreq_evtchn); + spin_unlock(&s->bufioreq_lock); + + return IOREQ_STATUS_HANDLED; +} + +int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, + bool buffered) +{ + struct vcpu *curr = current; + struct domain *d = curr->domain; + struct hvm_ioreq_vcpu *sv; + + ASSERT(s); + + if ( buffered ) + return hvm_send_buffered_ioreq(s, proto_p); + + if ( unlikely(!vcpu_start_shutdown_deferral(curr)) ) + return IOREQ_STATUS_RETRY; + + list_for_each_entry ( sv, + &s->ioreq_vcpu_list, + list_entry ) + { + if ( sv->vcpu == curr ) + { + evtchn_port_t port = sv->ioreq_evtchn; + ioreq_t *p = get_ioreq(s, curr); + + if ( unlikely(p->state != STATE_IOREQ_NONE) ) + { + gprintk(XENLOG_ERR, "device model set bad IO state %d\n", + p->state); + break; + } + + if ( unlikely(p->vp_eport != port) ) + { + gprintk(XENLOG_ERR, "device model set bad event channel %d\n", + p->vp_eport); + break; + } + + proto_p->state = STATE_IOREQ_NONE; + proto_p->vp_eport = port; + *p = *proto_p; + + prepare_wait_on_xen_event_channel(port); + + /* + * Following happens /after/ blocking and setting up ioreq + * contents. prepare_wait_on_xen_event_channel() is an implicit + * barrier. + */ + p->state = STATE_IOREQ_READY; + notify_via_xen_event_channel(d, port); + + sv->pending = true; + return IOREQ_STATUS_RETRY; + } + } + + return IOREQ_STATUS_UNHANDLED; +} + +unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) +{ + struct domain *d = current->domain; + struct hvm_ioreq_server *s; + unsigned int id, failed = 0; + + FOR_EACH_IOREQ_SERVER(d, id, s) + { + if ( !s->enabled ) + continue; + + if ( hvm_send_ioreq(s, p, buffered) == IOREQ_STATUS_UNHANDLED ) + failed++; + } + + return failed; +} + +void hvm_ioreq_init(struct domain *d) +{ + spin_lock_init(&d->arch.hvm.ioreq_server.lock); + + arch_ioreq_domain_init(d); +} + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/include/asm-x86/hvm/ioreq.h b/xen/include/asm-x86/hvm/ioreq.h index df0c292..9b2eb6f 100644 --- a/xen/include/asm-x86/hvm/ioreq.h +++ b/xen/include/asm-x86/hvm/ioreq.h @@ -19,42 +19,6 @@ #ifndef __ASM_X86_HVM_IOREQ_H__ #define __ASM_X86_HVM_IOREQ_H__ -bool hvm_io_pending(struct vcpu *v); -bool handle_hvm_io_completion(struct vcpu *v); -bool is_ioreq_server_page(struct domain *d, const struct page_info *page); - -int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, - ioservid_t *id); -int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id); -int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, - unsigned long *ioreq_gfn, - unsigned long *bufioreq_gfn, - evtchn_port_t *bufioreq_port); -int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, - unsigned long idx, mfn_t *mfn); -int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end); -int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end); -int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint32_t flags); -int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, - bool enabled); - -int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v); -void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v); -void hvm_destroy_all_ioreq_servers(struct domain *d); - -struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p); -int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, - bool buffered); -unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); - -void hvm_ioreq_init(struct domain *d); - /* This correlation must not be altered */ #define IOREQ_STATUS_HANDLED X86EMUL_OKAY #define IOREQ_STATUS_UNHANDLED X86EMUL_UNHANDLEABLE diff --git a/xen/include/asm-x86/ioreq.h b/xen/include/asm-x86/ioreq.h new file mode 100644 index 0000000..2aed60c --- /dev/null +++ b/xen/include/asm-x86/ioreq.h @@ -0,0 +1,37 @@ +/* + * ioreq.h: Hardware virtual machine assist interface definitions. + * + * This is a wrapper which purpose is to not include arch HVM specific header + * from the common code. + * + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#ifndef __ASM_X86_IOREQ_H__ +#define __ASM_X86_IOREQ_H__ + +#include + +#endif /* __ASM_X86_IOREQ_H__ */ + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index b95d3ef..430fc22 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -21,9 +21,47 @@ #include +#include + #define HANDLE_BUFIOREQ(s) \ ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) +bool hvm_io_pending(struct vcpu *v); +bool handle_hvm_io_completion(struct vcpu *v); +bool is_ioreq_server_page(struct domain *d, const struct page_info *page); + +int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, + ioservid_t *id); +int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id); +int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, + unsigned long *ioreq_gfn, + unsigned long *bufioreq_gfn, + evtchn_port_t *bufioreq_port); +int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, + unsigned long idx, mfn_t *mfn); +int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint64_t start, + uint64_t end); +int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint64_t start, + uint64_t end); +int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, + uint32_t type, uint32_t flags); +int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, + bool enabled); + +int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v); +void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v); +void hvm_destroy_all_ioreq_servers(struct domain *d); + +struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, + ioreq_t *p); +int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, + bool buffered); +unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); + +void hvm_ioreq_init(struct domain *d); + bool arch_ioreq_complete_mmio(void); bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s); From patchwork Mon Jan 25 19:08:12 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044039 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id CE4AAC433DB for ; Mon, 25 Jan 2021 19:09:24 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 916692067B for ; Mon, 25 Jan 2021 19:09:24 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 916692067B Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74300.133549 (Exim 4.92) (envelope-from ) id 1l47Ei-0000Dn-Tc; Mon, 25 Jan 2021 19:09:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74300.133549; Mon, 25 Jan 2021 19:09:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ei-0000Dc-PM; Mon, 25 Jan 2021 19:09:16 +0000 Received: by outflank-mailman (input) for mailman id 74300; Mon, 25 Jan 2021 19:09:15 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Eh-0008N7-GQ for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:15 +0000 Received: from mail-wm1-x32f.google.com (unknown [2a00:1450:4864:20::32f]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 9bfbffa9-99bf-4ce0-b91f-f14539646bd1; Mon, 25 Jan 2021 19:08:58 +0000 (UTC) Received: by mail-wm1-x32f.google.com with SMTP id c128so545130wme.2 for ; Mon, 25 Jan 2021 11:08:58 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.56 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:57 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 9bfbffa9-99bf-4ce0-b91f-f14539646bd1 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=xxdxnb18IQZyQw5ZgPoetsUdOChDIHKtZYMdcL56th8=; b=FsIzJwtHTH9vW8AnlkAXm3RWhRKHKLoV3ChKDtQIyyVp9HApSkaPDBbVf+22GvQUXS r0STc8tkCt8xceA4kpvHyuN+wDv4BOvnYUEqCq6rWN0JDOhS0ACrB+SJdzjdgT5KLAdz YLeFVTIJMsxgZorparLXgB4uz5EZ3n1qI/iBTv0U5mvEPjh8e+OPtXDoD4meLPz9PpUe WFLKtWsTyQFrrMvnpNWC20Ev6Hys/VeLLAD9mUXHXR+Q8yu+4Rz48YwJHaT2rRSQTeK9 hZaDdw90kP1B66BjPsPWj2xDt0Fmt+rYGg8NJaPzZ0BL3pVTbHsILYRZmeMMJBhIp0+k kx7g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=xxdxnb18IQZyQw5ZgPoetsUdOChDIHKtZYMdcL56th8=; b=HpbGx1oM0LGYizOqDnCBP+LIuAPeAjMNwf4hDrIJfpLHkNtGsaRh0WCvJSPbpY5L0L ijoa3bdRY4g/yrkr1bc0I1WzsJl73IuNr4MzlugZQ7MFucYr00HjDbc5rZJNe1To+VBE DK4w6jWjxy5/ymj2/TaECmia1vkuTtN7YVlCZie6YB7TowZETPSIGypo/J/g+82qlIEv xCkp5lXXF0zLadIjD8WILb+13/UAuIFnlTzETsA8ChKdOIpPPGc5zVzQSBEk1x2+yGS+ k+5iSn85tDtyU1mQ3YNyvHAMegAVHrWCE63qP0XaCLZlJouyDBsr+OnG0WYe4yjK+6ff JS9A== X-Gm-Message-State: AOAM532A1avjb9LUV/vHI/G5Hw6Q46FPI8chWr7HAeNsMuNlFKIql4cJ WxyQ7q0s+0d4XG5N85mgRieYdk9o8W3CMg== X-Google-Smtp-Source: ABdhPJxB5/BXkVwijZ/Ht81JLw1VC6Ay7ZTXnKX9A4yseCn/W0p+qmB1wA/DCz69LvKD68FDmipp9A== X-Received: by 2002:a1c:398b:: with SMTP id g133mr1415480wma.35.1611601737842; Mon, 25 Jan 2021 11:08:57 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 05/22] xen/ioreq: Make x86's hvm_ioreq_needs_completion() common Date: Mon, 25 Jan 2021 21:08:12 +0200 Message-Id: <1611601709-28361-6-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko The IOREQ is a common feature now and this helper will be used on Arm as is. Move it to xen/ioreq.h and remove "hvm" prefix. Although PIO handling on Arm is not introduced with the current series (it will be implemented when we add support for vPCI), technically the PIOs exist on Arm (however they are accessed the same way as MMIO) and it would be better not to diverge now. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Paul Durrant Acked-by: Jan Beulich Reviewed-by: Julien Grall Reviewed-by: Alex Bennée CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch, was split from: "[RFC PATCH V1 01/12] hvm/ioreq: Make x86's IOREQ feature common" Changes V1 -> V2: - remove "hvm" prefix Changes V2 -> V3: - add Paul's R-b Changes V3 -> V4: - add Jan's A-b Changes V4 -> V5: - rebase - add Julien's and Alex's R-b --- --- xen/arch/x86/hvm/emulate.c | 4 ++-- xen/arch/x86/hvm/io.c | 2 +- xen/common/ioreq.c | 4 ++-- xen/include/asm-x86/hvm/vcpu.h | 7 ------- xen/include/xen/ioreq.h | 7 +++++++ 5 files changed, 12 insertions(+), 12 deletions(-) diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c index 60ca465..c3487b5 100644 --- a/xen/arch/x86/hvm/emulate.c +++ b/xen/arch/x86/hvm/emulate.c @@ -336,7 +336,7 @@ static int hvmemul_do_io( rc = hvm_send_ioreq(s, &p, 0); if ( rc != X86EMUL_RETRY || currd->is_shutting_down ) vio->io_req.state = STATE_IOREQ_NONE; - else if ( !hvm_ioreq_needs_completion(&vio->io_req) ) + else if ( !ioreq_needs_completion(&vio->io_req) ) rc = X86EMUL_OKAY; } break; @@ -2649,7 +2649,7 @@ static int _hvm_emulate_one(struct hvm_emulate_ctxt *hvmemul_ctxt, if ( rc == X86EMUL_OKAY && vio->mmio_retry ) rc = X86EMUL_RETRY; - if ( !hvm_ioreq_needs_completion(&vio->io_req) ) + if ( !ioreq_needs_completion(&vio->io_req) ) completion = HVMIO_no_completion; else if ( completion == HVMIO_no_completion ) completion = (vio->io_req.type != IOREQ_TYPE_PIO || diff --git a/xen/arch/x86/hvm/io.c b/xen/arch/x86/hvm/io.c index 11e007d..ef8286b 100644 --- a/xen/arch/x86/hvm/io.c +++ b/xen/arch/x86/hvm/io.c @@ -135,7 +135,7 @@ bool handle_pio(uint16_t port, unsigned int size, int dir) rc = hvmemul_do_pio_buffer(port, size, dir, &data); - if ( hvm_ioreq_needs_completion(&vio->io_req) ) + if ( ioreq_needs_completion(&vio->io_req) ) vio->io_completion = HVMIO_pio_completion; switch ( rc ) diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 4e7d91b..61ddd54 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -160,7 +160,7 @@ static bool hvm_wait_for_io(struct hvm_ioreq_vcpu *sv, ioreq_t *p) } p = &sv->vcpu->arch.hvm.hvm_io.io_req; - if ( hvm_ioreq_needs_completion(p) ) + if ( ioreq_needs_completion(p) ) p->data = data; sv->pending = false; @@ -186,7 +186,7 @@ bool handle_hvm_io_completion(struct vcpu *v) if ( sv && !hvm_wait_for_io(sv, get_ioreq(s, v)) ) return false; - vio->io_req.state = hvm_ioreq_needs_completion(&vio->io_req) ? + vio->io_req.state = ioreq_needs_completion(&vio->io_req) ? STATE_IORESP_READY : STATE_IOREQ_NONE; msix_write_completion(v); diff --git a/xen/include/asm-x86/hvm/vcpu.h b/xen/include/asm-x86/hvm/vcpu.h index 5ccd075..6c1feda 100644 --- a/xen/include/asm-x86/hvm/vcpu.h +++ b/xen/include/asm-x86/hvm/vcpu.h @@ -91,13 +91,6 @@ struct hvm_vcpu_io { const struct g2m_ioport *g2m_ioport; }; -static inline bool hvm_ioreq_needs_completion(const ioreq_t *ioreq) -{ - return ioreq->state == STATE_IOREQ_READY && - !ioreq->data_is_ptr && - (ioreq->type != IOREQ_TYPE_PIO || ioreq->dir != IOREQ_WRITE); -} - struct nestedvcpu { bool_t nv_guestmode; /* vcpu in guestmode? */ void *nv_vvmcx; /* l1 guest virtual VMCB/VMCS */ diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 430fc22..e957b52 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -23,6 +23,13 @@ #include +static inline bool ioreq_needs_completion(const ioreq_t *ioreq) +{ + return ioreq->state == STATE_IOREQ_READY && + !ioreq->data_is_ptr && + (ioreq->type != IOREQ_TYPE_PIO || ioreq->dir != IOREQ_WRITE); +} + #define HANDLE_BUFIOREQ(s) \ ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) From patchwork Mon Jan 25 19:08:13 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044041 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 58FFCC433DB for ; Mon, 25 Jan 2021 19:09:31 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 0F04F2067B for ; Mon, 25 Jan 2021 19:09:31 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 0F04F2067B Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74301.133561 (Exim 4.92) (envelope-from ) id 1l47Eo-0000JA-8K; Mon, 25 Jan 2021 19:09:22 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74301.133561; Mon, 25 Jan 2021 19:09:22 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Eo-0000Iy-3P; Mon, 25 Jan 2021 19:09:22 +0000 Received: by outflank-mailman (input) for mailman id 74301; Mon, 25 Jan 2021 19:09:20 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Em-0008N7-GT for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:20 +0000 Received: from mail-wr1-x42a.google.com (unknown [2a00:1450:4864:20::42a]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id b347190a-9f79-4f28-9d72-0066c8359a55; Mon, 25 Jan 2021 19:08:59 +0000 (UTC) Received: by mail-wr1-x42a.google.com with SMTP id l12so14101918wry.2 for ; Mon, 25 Jan 2021 11:08:59 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.57 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:58 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: b347190a-9f79-4f28-9d72-0066c8359a55 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=BoY0i+xc8eGVnOTYuk/xVaKxY5o/2NyqxPbuZ81MBYw=; b=s1nla7VbneAaYxYlqA3tdXVAWfBiEpsHuGdiKzAwNTq5bna17/7bD2ZtSNUcdRy8yr NbE0SQVcbT4WUe6jxii3Vs8PKrQ5FzpvTFprqw1jzZSkwD27FfVXe+UmoNTkA1AD3SPl PvAqFfHqC/Ru0uw+vSkZv4TAi8NpAXNxFXmJNReGMFrUPjtNDi5G0F8mSPCT9LAdQpcb mJHANrNo6/xZcKzxZEFG265rbB7EtC0EMCzd1rwKeBfEOVDh9h1s3s66Hi92M7k+q4d/ HRt5UmUunhmJ7QLj/j0tC268oh50D22Gbrw5nir80ieqKBg9tj8K6+cKr5XE6wA7OyED D4cA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=BoY0i+xc8eGVnOTYuk/xVaKxY5o/2NyqxPbuZ81MBYw=; b=W/NSNuIdj7KtOC8T1ntvh8CP0pj5ZHgZoGXbu3QRVhy3NWXZUPk9ylw0Y4BjsxPYOi tb2N6uzygTOgsdjlUMfbYGq8pkI8R/4dmQh07EMbHXtWcsH1Nb0P2nkl4P2SZpHDiZRi uukgrpq2msRma/5BAZ+Nfe6zib9/edY23YhLCdrJN/xLh3z5Qy9c5GFA+O6/vTmdPU9I 7KOd2tfw/CphO2BHm9jekktzc38iR66udytY7SkIenwTI2NjclAAanmtxLm84J+eFmM4 NazyVAKVO2T8jbMLw4CaFv/kShCVCDsr/ezzIjp7JtlXnfJ+BzmN8oXZ1LZh1y8LXDhg /F/A== X-Gm-Message-State: AOAM530SaKKncY+BQaeLxBNAJbMA9v39DmqdwNRbMAkv6B2zhCygG1i2 V+IM38HhtruqGzgbh/6YF0c886KVDH0fpg== X-Google-Smtp-Source: ABdhPJyiSLRJ6NSIB7W89yFmbKIMy676yPxZocuTupN4lW46MeIWve2sPqz817trNF65a6GbuwuNpg== X-Received: by 2002:a5d:452f:: with SMTP id j15mr2598185wra.298.1611601738986; Mon, 25 Jan 2021 11:08:58 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 06/22] xen/ioreq: Make x86's hvm_mmio_first(last)_byte() common Date: Mon, 25 Jan 2021 21:08:13 +0200 Message-Id: <1611601709-28361-7-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko The IOREQ is a common feature now and these helpers will be used on Arm as is. Move them to xen/ioreq.h and replace "hvm" prefixes with "ioreq". Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Paul Durrant Acked-by: Jan Beulich Reviewed-by: Julien Grall Reviewed-by: Alex Bennée CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch Changes V1 -> V2: - replace "hvm" prefix by "ioreq" Changes V2 -> V3: - add Paul's R-b Changes V3 -> V4: - add Jan's A-b Changes V4 -> V5: - rebase - add Julien's and Alex's R-b --- --- xen/arch/x86/hvm/intercept.c | 5 +++-- xen/arch/x86/hvm/stdvga.c | 4 ++-- xen/common/ioreq.c | 4 ++-- xen/include/asm-x86/hvm/io.h | 16 ---------------- xen/include/xen/ioreq.h | 16 ++++++++++++++++ 5 files changed, 23 insertions(+), 22 deletions(-) diff --git a/xen/arch/x86/hvm/intercept.c b/xen/arch/x86/hvm/intercept.c index cd4c4c1..02ca3b0 100644 --- a/xen/arch/x86/hvm/intercept.c +++ b/xen/arch/x86/hvm/intercept.c @@ -17,6 +17,7 @@ * this program; If not, see . */ +#include #include #include #include @@ -34,7 +35,7 @@ static bool_t hvm_mmio_accept(const struct hvm_io_handler *handler, const ioreq_t *p) { - paddr_t first = hvm_mmio_first_byte(p), last; + paddr_t first = ioreq_mmio_first_byte(p), last; BUG_ON(handler->type != IOREQ_TYPE_COPY); @@ -42,7 +43,7 @@ static bool_t hvm_mmio_accept(const struct hvm_io_handler *handler, return 0; /* Make sure the handler will accept the whole access. */ - last = hvm_mmio_last_byte(p); + last = ioreq_mmio_last_byte(p); if ( last != first && !handler->mmio.ops->check(current, last) ) domain_crash(current->domain); diff --git a/xen/arch/x86/hvm/stdvga.c b/xen/arch/x86/hvm/stdvga.c index fd7cadb..17dee74 100644 --- a/xen/arch/x86/hvm/stdvga.c +++ b/xen/arch/x86/hvm/stdvga.c @@ -524,8 +524,8 @@ static bool_t stdvga_mem_accept(const struct hvm_io_handler *handler, * deadlock when hvm_mmio_internal() is called from * hvm_copy_to/from_guest_phys() in hvm_process_io_intercept(). */ - if ( (hvm_mmio_first_byte(p) < VGA_MEM_BASE) || - (hvm_mmio_last_byte(p) >= (VGA_MEM_BASE + VGA_MEM_SIZE)) ) + if ( (ioreq_mmio_first_byte(p) < VGA_MEM_BASE) || + (ioreq_mmio_last_byte(p) >= (VGA_MEM_BASE + VGA_MEM_SIZE)) ) return 0; spin_lock(&s->lock); diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 61ddd54..89e75ff 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -1078,8 +1078,8 @@ struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, break; case XEN_DMOP_IO_RANGE_MEMORY: - start = hvm_mmio_first_byte(p); - end = hvm_mmio_last_byte(p); + start = ioreq_mmio_first_byte(p); + end = ioreq_mmio_last_byte(p); if ( rangeset_contains_range(r, start, end) ) return s; diff --git a/xen/include/asm-x86/hvm/io.h b/xen/include/asm-x86/hvm/io.h index 9453b9b..6bc80db 100644 --- a/xen/include/asm-x86/hvm/io.h +++ b/xen/include/asm-x86/hvm/io.h @@ -40,22 +40,6 @@ struct hvm_mmio_ops { hvm_mmio_write_t write; }; -static inline paddr_t hvm_mmio_first_byte(const ioreq_t *p) -{ - return unlikely(p->df) ? - p->addr - (p->count - 1ul) * p->size : - p->addr; -} - -static inline paddr_t hvm_mmio_last_byte(const ioreq_t *p) -{ - unsigned long size = p->size; - - return unlikely(p->df) ? - p->addr + size - 1: - p->addr + (p->count * size) - 1; -} - typedef int (*portio_action_t)( int dir, unsigned int port, unsigned int bytes, uint32_t *val); diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index e957b52..6853aa3 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -23,6 +23,22 @@ #include +static inline paddr_t ioreq_mmio_first_byte(const ioreq_t *p) +{ + return unlikely(p->df) ? + p->addr - (p->count - 1ul) * p->size : + p->addr; +} + +static inline paddr_t ioreq_mmio_last_byte(const ioreq_t *p) +{ + unsigned long size = p->size; + + return unlikely(p->df) ? + p->addr + size - 1: + p->addr + (p->count * size) - 1; +} + static inline bool ioreq_needs_completion(const ioreq_t *ioreq) { return ioreq->state == STATE_IOREQ_READY && From patchwork Mon Jan 25 19:08:14 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044043 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 39E2BC433E0 for ; Mon, 25 Jan 2021 19:09:39 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id E51822067B for ; Mon, 25 Jan 2021 19:09:38 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org E51822067B Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74305.133585 (Exim 4.92) (envelope-from ) id 1l47Ex-0000Tb-4A; Mon, 25 Jan 2021 19:09:31 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74305.133585; Mon, 25 Jan 2021 19:09:31 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ex-0000TO-04; Mon, 25 Jan 2021 19:09:31 +0000 Received: by outflank-mailman (input) for mailman id 74305; Mon, 25 Jan 2021 19:09:30 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ew-0008N7-GX for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:30 +0000 Received: from mail-wr1-x42e.google.com (unknown [2a00:1450:4864:20::42e]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 35a65785-2f20-4a8b-b91a-7c2f6b2527f6; Mon, 25 Jan 2021 19:09:01 +0000 (UTC) Received: by mail-wr1-x42e.google.com with SMTP id l12so14101970wry.2 for ; Mon, 25 Jan 2021 11:09:01 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.08.59 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:08:59 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 35a65785-2f20-4a8b-b91a-7c2f6b2527f6 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=cLeMDNEt5nkku7uIZlyHSW2fNYGg+Q0P2/YAJRwHIeU=; b=gmVpAmebCj2zsry46JmQmaQCNRvdzN+I78/n9JseUFDeACrCInzy/UfducW9rHhdBt ZWSVdKZRlodjneJWHnXVLLkDiZrTuoLjwXOjqxH5KMc8jdfyVpCYYcxAXRQWpcuT4+YW E/skkm/R3TR9A+AAoAJKrTQ/XXGyIrZGG1XHPRJWA4cgBNFXveGIhsgPoa8fdVLyIWBJ jj5gv1NxahXWqf5t5HUQf3nhsnhNbgj3vsCgLrZfmmotnGOcUAXSu8cW7bOcJBsEnRHI 0Wu9cEK/4/3WbU6h6erH/8bot5g3Hxg6VVIBZzQwTZ6To/cnuPpGGLovEFXUVucXiFns k1TQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=cLeMDNEt5nkku7uIZlyHSW2fNYGg+Q0P2/YAJRwHIeU=; b=b5FcwgDG4I21x02BbneAM1IOh1hCms9crRDx1YDrsAAF4TYFcFGiaPAGyTwBlvPBuK zJUWlG6bXtC4Xba7dYrfy/85jjheg0Ete3KYnHBvLoRLmTpNeqEh9qbSyBG7RXYh9RIh +ad34suZ8k+UhPuCxOMLa8v5Tbq8gUbCINFlUSrJYYM9dp33LFiXuV9iMnYO47kZKbQ3 39+9dPkdzlUv/R7sLbFzbeVLgloLSILhzLii3G0rIyJUvepAheC6SlAkjzlVOgfjUkbP W7My12CNpmw60E5SVjBWgnNOkupUbIQv+EID+E5CchIx4jCdrBFQmordCO3Tq1Pe1d6x Slyw== X-Gm-Message-State: AOAM530uiVi3kDRLqNfYRfKcSX+hKd5ZLvylbrfmzv4JnLCJeqBVVypL CdziE66CcREgSXYSzGzOHbKugNKWwWFpzg== X-Google-Smtp-Source: ABdhPJyKQGPA75urwn0gSZPiT/aOmX7xS1TzTe6+6A1YhXhfQxqRnh8TfwQA0swgiWIE+BXpEkfwCg== X-Received: by 2002:adf:e4c9:: with SMTP id v9mr2491108wrm.277.1611601740206; Mon, 25 Jan 2021 11:09:00 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 07/22] xen/ioreq: Make x86's hvm_ioreq_(page/vcpu/server) structs common Date: Mon, 25 Jan 2021 21:08:14 +0200 Message-Id: <1611601709-28361-8-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko The IOREQ is a common feature now and these structs will be used on Arm as is. Move them to xen/ioreq.h and remove "hvm" prefixes. Signed-off-by: Oleksandr Tyshchenko Acked-by: Jan Beulich Reviewed-by: Julien Grall Reviewed-by: Paul Durrant Reviewed-by: Alex Bennée CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch Changes V1 -> V2: - remove "hvm" prefix Changes V2 -> V3: - update patch according the "legacy interface" is x86 specific Changes V3 -> V4: - add Jan's A-b Changes V4 -> V5: - rebase - add Julien's, Alex's and Paul's R-b - fix alignment issue in domain.h --- --- xen/arch/x86/hvm/emulate.c | 2 +- xen/arch/x86/hvm/ioreq.c | 38 +++++++------- xen/arch/x86/hvm/stdvga.c | 2 +- xen/arch/x86/mm/p2m.c | 8 +-- xen/common/ioreq.c | 108 +++++++++++++++++++-------------------- xen/include/asm-x86/hvm/domain.h | 36 +------------ xen/include/asm-x86/p2m.h | 8 +-- xen/include/xen/ioreq.h | 54 ++++++++++++++++---- 8 files changed, 128 insertions(+), 128 deletions(-) diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c index c3487b5..4d62199 100644 --- a/xen/arch/x86/hvm/emulate.c +++ b/xen/arch/x86/hvm/emulate.c @@ -287,7 +287,7 @@ static int hvmemul_do_io( * However, there's no cheap approach to avoid above situations in xen, * so the device model side needs to check the incoming ioreq event. */ - struct hvm_ioreq_server *s = NULL; + struct ioreq_server *s = NULL; p2m_type_t p2mt = p2m_invalid; if ( is_mmio ) diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 666d695..0cadf34 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -63,7 +63,7 @@ bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion) return true; } -static gfn_t hvm_alloc_legacy_ioreq_gfn(struct hvm_ioreq_server *s) +static gfn_t hvm_alloc_legacy_ioreq_gfn(struct ioreq_server *s) { struct domain *d = s->target; unsigned int i; @@ -79,7 +79,7 @@ static gfn_t hvm_alloc_legacy_ioreq_gfn(struct hvm_ioreq_server *s) return INVALID_GFN; } -static gfn_t hvm_alloc_ioreq_gfn(struct hvm_ioreq_server *s) +static gfn_t hvm_alloc_ioreq_gfn(struct ioreq_server *s) { struct domain *d = s->target; unsigned int i; @@ -97,7 +97,7 @@ static gfn_t hvm_alloc_ioreq_gfn(struct hvm_ioreq_server *s) return hvm_alloc_legacy_ioreq_gfn(s); } -static bool hvm_free_legacy_ioreq_gfn(struct hvm_ioreq_server *s, +static bool hvm_free_legacy_ioreq_gfn(struct ioreq_server *s, gfn_t gfn) { struct domain *d = s->target; @@ -115,7 +115,7 @@ static bool hvm_free_legacy_ioreq_gfn(struct hvm_ioreq_server *s, return true; } -static void hvm_free_ioreq_gfn(struct hvm_ioreq_server *s, gfn_t gfn) +static void hvm_free_ioreq_gfn(struct ioreq_server *s, gfn_t gfn) { struct domain *d = s->target; unsigned int i = gfn_x(gfn) - d->arch.hvm.ioreq_gfn.base; @@ -129,9 +129,9 @@ static void hvm_free_ioreq_gfn(struct hvm_ioreq_server *s, gfn_t gfn) } } -static void hvm_unmap_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) +static void hvm_unmap_ioreq_gfn(struct ioreq_server *s, bool buf) { - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; if ( gfn_eq(iorp->gfn, INVALID_GFN) ) return; @@ -143,10 +143,10 @@ static void hvm_unmap_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) iorp->gfn = INVALID_GFN; } -static int hvm_map_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) +static int hvm_map_ioreq_gfn(struct ioreq_server *s, bool buf) { struct domain *d = s->target; - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; int rc; if ( iorp->page ) @@ -179,11 +179,11 @@ static int hvm_map_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) return rc; } -static void hvm_remove_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) +static void hvm_remove_ioreq_gfn(struct ioreq_server *s, bool buf) { struct domain *d = s->target; - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; if ( gfn_eq(iorp->gfn, INVALID_GFN) ) return; @@ -194,10 +194,10 @@ static void hvm_remove_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) clear_page(iorp->va); } -static int hvm_add_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) +static int hvm_add_ioreq_gfn(struct ioreq_server *s, bool buf) { struct domain *d = s->target; - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; int rc; if ( gfn_eq(iorp->gfn, INVALID_GFN) ) @@ -213,7 +213,7 @@ static int hvm_add_ioreq_gfn(struct hvm_ioreq_server *s, bool buf) return rc; } -int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s) +int arch_ioreq_server_map_pages(struct ioreq_server *s) { int rc; @@ -228,40 +228,40 @@ int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s) return rc; } -void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s) +void arch_ioreq_server_unmap_pages(struct ioreq_server *s) { hvm_unmap_ioreq_gfn(s, true); hvm_unmap_ioreq_gfn(s, false); } -void arch_ioreq_server_enable(struct hvm_ioreq_server *s) +void arch_ioreq_server_enable(struct ioreq_server *s) { hvm_remove_ioreq_gfn(s, false); hvm_remove_ioreq_gfn(s, true); } -void arch_ioreq_server_disable(struct hvm_ioreq_server *s) +void arch_ioreq_server_disable(struct ioreq_server *s) { hvm_add_ioreq_gfn(s, true); hvm_add_ioreq_gfn(s, false); } /* Called when target domain is paused */ -void arch_ioreq_server_destroy(struct hvm_ioreq_server *s) +void arch_ioreq_server_destroy(struct ioreq_server *s) { p2m_set_ioreq_server(s->target, 0, s); } /* Called with ioreq_server lock held */ int arch_ioreq_server_map_mem_type(struct domain *d, - struct hvm_ioreq_server *s, + struct ioreq_server *s, uint32_t flags) { return p2m_set_ioreq_server(d, flags, s); } void arch_ioreq_server_map_mem_type_completed(struct domain *d, - struct hvm_ioreq_server *s, + struct ioreq_server *s, uint32_t flags) { if ( flags == 0 && read_atomic(&p2m_get_hostp2m(d)->ioreq.entry_count) ) diff --git a/xen/arch/x86/hvm/stdvga.c b/xen/arch/x86/hvm/stdvga.c index 17dee74..ee13449 100644 --- a/xen/arch/x86/hvm/stdvga.c +++ b/xen/arch/x86/hvm/stdvga.c @@ -466,7 +466,7 @@ static int stdvga_mem_write(const struct hvm_io_handler *handler, .dir = IOREQ_WRITE, .data = data, }; - struct hvm_ioreq_server *srv; + struct ioreq_server *srv; if ( !stdvga_cache_is_enabled(s) || !s->stdvga ) goto done; diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c index a32301c..c1dd45b 100644 --- a/xen/arch/x86/mm/p2m.c +++ b/xen/arch/x86/mm/p2m.c @@ -372,7 +372,7 @@ void p2m_memory_type_changed(struct domain *d) int p2m_set_ioreq_server(struct domain *d, unsigned int flags, - struct hvm_ioreq_server *s) + struct ioreq_server *s) { struct p2m_domain *p2m = p2m_get_hostp2m(d); int rc; @@ -420,11 +420,11 @@ int p2m_set_ioreq_server(struct domain *d, return rc; } -struct hvm_ioreq_server *p2m_get_ioreq_server(struct domain *d, - unsigned int *flags) +struct ioreq_server *p2m_get_ioreq_server(struct domain *d, + unsigned int *flags) { struct p2m_domain *p2m = p2m_get_hostp2m(d); - struct hvm_ioreq_server *s; + struct ioreq_server *s; spin_lock(&p2m->ioreq.lock); diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 89e75ff..7320f23 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -35,7 +35,7 @@ #include static void set_ioreq_server(struct domain *d, unsigned int id, - struct hvm_ioreq_server *s) + struct ioreq_server *s) { ASSERT(id < MAX_NR_IOREQ_SERVERS); ASSERT(!s || !d->arch.hvm.ioreq_server.server[id]); @@ -46,8 +46,8 @@ static void set_ioreq_server(struct domain *d, unsigned int id, #define GET_IOREQ_SERVER(d, id) \ (d)->arch.hvm.ioreq_server.server[id] -static struct hvm_ioreq_server *get_ioreq_server(const struct domain *d, - unsigned int id) +static struct ioreq_server *get_ioreq_server(const struct domain *d, + unsigned int id) { if ( id >= MAX_NR_IOREQ_SERVERS ) return NULL; @@ -69,7 +69,7 @@ static struct hvm_ioreq_server *get_ioreq_server(const struct domain *d, continue; \ else -static ioreq_t *get_ioreq(struct hvm_ioreq_server *s, struct vcpu *v) +static ioreq_t *get_ioreq(struct ioreq_server *s, struct vcpu *v) { shared_iopage_t *p = s->ioreq.va; @@ -79,16 +79,16 @@ static ioreq_t *get_ioreq(struct hvm_ioreq_server *s, struct vcpu *v) return &p->vcpu_ioreq[v->vcpu_id]; } -static struct hvm_ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, - struct hvm_ioreq_server **srvp) +static struct ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, + struct ioreq_server **srvp) { struct domain *d = v->domain; - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int id; FOR_EACH_IOREQ_SERVER(d, id, s) { - struct hvm_ioreq_vcpu *sv; + struct ioreq_vcpu *sv; list_for_each_entry ( sv, &s->ioreq_vcpu_list, @@ -111,7 +111,7 @@ bool hvm_io_pending(struct vcpu *v) return get_pending_vcpu(v, NULL); } -static bool hvm_wait_for_io(struct hvm_ioreq_vcpu *sv, ioreq_t *p) +static bool hvm_wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) { unsigned int prev_state = STATE_IOREQ_NONE; unsigned int state = p->state; @@ -172,8 +172,8 @@ bool handle_hvm_io_completion(struct vcpu *v) { struct domain *d = v->domain; struct hvm_vcpu_io *vio = &v->arch.hvm.hvm_io; - struct hvm_ioreq_server *s; - struct hvm_ioreq_vcpu *sv; + struct ioreq_server *s; + struct ioreq_vcpu *sv; enum hvm_io_completion io_completion; if ( has_vpci(d) && vpci_process_pending(v) ) @@ -214,9 +214,9 @@ bool handle_hvm_io_completion(struct vcpu *v) return true; } -static int hvm_alloc_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) +static int hvm_alloc_ioreq_mfn(struct ioreq_server *s, bool buf) { - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; struct page_info *page; if ( iorp->page ) @@ -262,9 +262,9 @@ static int hvm_alloc_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) return -ENOMEM; } -static void hvm_free_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) +static void hvm_free_ioreq_mfn(struct ioreq_server *s, bool buf) { - struct hvm_ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; + struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; struct page_info *page = iorp->page; if ( !page ) @@ -281,7 +281,7 @@ static void hvm_free_ioreq_mfn(struct hvm_ioreq_server *s, bool buf) bool is_ioreq_server_page(struct domain *d, const struct page_info *page) { - const struct hvm_ioreq_server *s; + const struct ioreq_server *s; unsigned int id; bool found = false; @@ -301,8 +301,8 @@ bool is_ioreq_server_page(struct domain *d, const struct page_info *page) return found; } -static void hvm_update_ioreq_evtchn(struct hvm_ioreq_server *s, - struct hvm_ioreq_vcpu *sv) +static void hvm_update_ioreq_evtchn(struct ioreq_server *s, + struct ioreq_vcpu *sv) { ASSERT(spin_is_locked(&s->lock)); @@ -314,13 +314,13 @@ static void hvm_update_ioreq_evtchn(struct hvm_ioreq_server *s, } } -static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, +static int hvm_ioreq_server_add_vcpu(struct ioreq_server *s, struct vcpu *v) { - struct hvm_ioreq_vcpu *sv; + struct ioreq_vcpu *sv; int rc; - sv = xzalloc(struct hvm_ioreq_vcpu); + sv = xzalloc(struct ioreq_vcpu); rc = -ENOMEM; if ( !sv ) @@ -366,10 +366,10 @@ static int hvm_ioreq_server_add_vcpu(struct hvm_ioreq_server *s, return rc; } -static void hvm_ioreq_server_remove_vcpu(struct hvm_ioreq_server *s, +static void hvm_ioreq_server_remove_vcpu(struct ioreq_server *s, struct vcpu *v) { - struct hvm_ioreq_vcpu *sv; + struct ioreq_vcpu *sv; spin_lock(&s->lock); @@ -394,9 +394,9 @@ static void hvm_ioreq_server_remove_vcpu(struct hvm_ioreq_server *s, spin_unlock(&s->lock); } -static void hvm_ioreq_server_remove_all_vcpus(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_remove_all_vcpus(struct ioreq_server *s) { - struct hvm_ioreq_vcpu *sv, *next; + struct ioreq_vcpu *sv, *next; spin_lock(&s->lock); @@ -420,7 +420,7 @@ static void hvm_ioreq_server_remove_all_vcpus(struct hvm_ioreq_server *s) spin_unlock(&s->lock); } -static int hvm_ioreq_server_alloc_pages(struct hvm_ioreq_server *s) +static int hvm_ioreq_server_alloc_pages(struct ioreq_server *s) { int rc; @@ -435,13 +435,13 @@ static int hvm_ioreq_server_alloc_pages(struct hvm_ioreq_server *s) return rc; } -static void hvm_ioreq_server_free_pages(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_free_pages(struct ioreq_server *s) { hvm_free_ioreq_mfn(s, true); hvm_free_ioreq_mfn(s, false); } -static void hvm_ioreq_server_free_rangesets(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_free_rangesets(struct ioreq_server *s) { unsigned int i; @@ -449,7 +449,7 @@ static void hvm_ioreq_server_free_rangesets(struct hvm_ioreq_server *s) rangeset_destroy(s->range[i]); } -static int hvm_ioreq_server_alloc_rangesets(struct hvm_ioreq_server *s, +static int hvm_ioreq_server_alloc_rangesets(struct ioreq_server *s, ioservid_t id) { unsigned int i; @@ -487,9 +487,9 @@ static int hvm_ioreq_server_alloc_rangesets(struct hvm_ioreq_server *s, return rc; } -static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_enable(struct ioreq_server *s) { - struct hvm_ioreq_vcpu *sv; + struct ioreq_vcpu *sv; spin_lock(&s->lock); @@ -509,7 +509,7 @@ static void hvm_ioreq_server_enable(struct hvm_ioreq_server *s) spin_unlock(&s->lock); } -static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_disable(struct ioreq_server *s) { spin_lock(&s->lock); @@ -524,7 +524,7 @@ static void hvm_ioreq_server_disable(struct hvm_ioreq_server *s) spin_unlock(&s->lock); } -static int hvm_ioreq_server_init(struct hvm_ioreq_server *s, +static int hvm_ioreq_server_init(struct ioreq_server *s, struct domain *d, int bufioreq_handling, ioservid_t id) { @@ -569,7 +569,7 @@ static int hvm_ioreq_server_init(struct hvm_ioreq_server *s, return rc; } -static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) +static void hvm_ioreq_server_deinit(struct ioreq_server *s) { ASSERT(!s->enabled); hvm_ioreq_server_remove_all_vcpus(s); @@ -594,14 +594,14 @@ static void hvm_ioreq_server_deinit(struct hvm_ioreq_server *s) int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, ioservid_t *id) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int i; int rc; if ( bufioreq_handling > HVM_IOREQSRV_BUFIOREQ_ATOMIC ) return -EINVAL; - s = xzalloc(struct hvm_ioreq_server); + s = xzalloc(struct ioreq_server); if ( !s ) return -ENOMEM; @@ -649,7 +649,7 @@ int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; int rc; spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); @@ -694,7 +694,7 @@ int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, unsigned long *bufioreq_gfn, evtchn_port_t *bufioreq_port) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; int rc; spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); @@ -739,7 +739,7 @@ int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, unsigned long idx, mfn_t *mfn) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; int rc; ASSERT(is_hvm_domain(d)); @@ -791,7 +791,7 @@ int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, uint32_t type, uint64_t start, uint64_t end) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; struct rangeset *r; int rc; @@ -843,7 +843,7 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, uint32_t type, uint64_t start, uint64_t end) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; struct rangeset *r; int rc; @@ -902,7 +902,7 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, uint32_t type, uint32_t flags) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; int rc; if ( type != HVMMEM_ioreq_server ) @@ -937,7 +937,7 @@ int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, bool enabled) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; int rc; spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); @@ -970,7 +970,7 @@ int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int id; int rc; @@ -1005,7 +1005,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int id; spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); @@ -1018,7 +1018,7 @@ void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) void hvm_destroy_all_ioreq_servers(struct domain *d) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int id; if ( !arch_ioreq_server_destroy_all(d) ) @@ -1045,10 +1045,10 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); } -struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p) +struct ioreq_server *hvm_select_ioreq_server(struct domain *d, + ioreq_t *p) { - struct hvm_ioreq_server *s; + struct ioreq_server *s; uint8_t type; uint64_t addr; unsigned int id; @@ -1101,10 +1101,10 @@ struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, return NULL; } -static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) +static int hvm_send_buffered_ioreq(struct ioreq_server *s, ioreq_t *p) { struct domain *d = current->domain; - struct hvm_ioreq_page *iorp; + struct ioreq_page *iorp; buffered_iopage_t *pg; buf_ioreq_t bp = { .data = p->data, .addr = p->addr, @@ -1194,12 +1194,12 @@ static int hvm_send_buffered_ioreq(struct hvm_ioreq_server *s, ioreq_t *p) return IOREQ_STATUS_HANDLED; } -int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, +int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, bool buffered) { struct vcpu *curr = current; struct domain *d = curr->domain; - struct hvm_ioreq_vcpu *sv; + struct ioreq_vcpu *sv; ASSERT(s); @@ -1257,7 +1257,7 @@ int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) { struct domain *d = current->domain; - struct hvm_ioreq_server *s; + struct ioreq_server *s; unsigned int id, failed = 0; FOR_EACH_IOREQ_SERVER(d, id, s) diff --git a/xen/include/asm-x86/hvm/domain.h b/xen/include/asm-x86/hvm/domain.h index 9d247ba..3b36c2f 100644 --- a/xen/include/asm-x86/hvm/domain.h +++ b/xen/include/asm-x86/hvm/domain.h @@ -30,40 +30,6 @@ #include -struct hvm_ioreq_page { - gfn_t gfn; - struct page_info *page; - void *va; -}; - -struct hvm_ioreq_vcpu { - struct list_head list_entry; - struct vcpu *vcpu; - evtchn_port_t ioreq_evtchn; - bool pending; -}; - -#define NR_IO_RANGE_TYPES (XEN_DMOP_IO_RANGE_PCI + 1) -#define MAX_NR_IO_RANGES 256 - -struct hvm_ioreq_server { - struct domain *target, *emulator; - - /* Lock to serialize toolstack modifications */ - spinlock_t lock; - - struct hvm_ioreq_page ioreq; - struct list_head ioreq_vcpu_list; - struct hvm_ioreq_page bufioreq; - - /* Lock to serialize access to buffered ioreq ring */ - spinlock_t bufioreq_lock; - evtchn_port_t bufioreq_evtchn; - struct rangeset *range[NR_IO_RANGE_TYPES]; - bool enabled; - uint8_t bufioreq_handling; -}; - #ifdef CONFIG_MEM_SHARING struct mem_sharing_domain { @@ -110,7 +76,7 @@ struct hvm_domain { /* Lock protects all other values in the sub-struct and the default */ struct { spinlock_t lock; - struct hvm_ioreq_server *server[MAX_NR_IOREQ_SERVERS]; + struct ioreq_server *server[MAX_NR_IOREQ_SERVERS]; } ioreq_server; /* Cached CF8 for guest PCI config cycles */ diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h index 6acbd1e..5d7836d 100644 --- a/xen/include/asm-x86/p2m.h +++ b/xen/include/asm-x86/p2m.h @@ -363,7 +363,7 @@ struct p2m_domain { * ioreq server who's responsible for the emulation of * gfns with specific p2m type(for now, p2m_ioreq_server). */ - struct hvm_ioreq_server *server; + struct ioreq_server *server; /* * flags specifies whether read, write or both operations * are to be emulated by an ioreq server. @@ -937,9 +937,9 @@ static inline unsigned int p2m_get_iommu_flags(p2m_type_t p2mt, mfn_t mfn) } int p2m_set_ioreq_server(struct domain *d, unsigned int flags, - struct hvm_ioreq_server *s); -struct hvm_ioreq_server *p2m_get_ioreq_server(struct domain *d, - unsigned int *flags); + struct ioreq_server *s); +struct ioreq_server *p2m_get_ioreq_server(struct domain *d, + unsigned int *flags); static inline int p2m_entry_modify(struct p2m_domain *p2m, p2m_type_t nt, p2m_type_t ot, mfn_t nfn, mfn_t ofn, diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 6853aa3..5a6c11d 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -23,6 +23,40 @@ #include +struct ioreq_page { + gfn_t gfn; + struct page_info *page; + void *va; +}; + +struct ioreq_vcpu { + struct list_head list_entry; + struct vcpu *vcpu; + evtchn_port_t ioreq_evtchn; + bool pending; +}; + +#define NR_IO_RANGE_TYPES (XEN_DMOP_IO_RANGE_PCI + 1) +#define MAX_NR_IO_RANGES 256 + +struct ioreq_server { + struct domain *target, *emulator; + + /* Lock to serialize toolstack modifications */ + spinlock_t lock; + + struct ioreq_page ioreq; + struct list_head ioreq_vcpu_list; + struct ioreq_page bufioreq; + + /* Lock to serialize access to buffered ioreq ring */ + spinlock_t bufioreq_lock; + evtchn_port_t bufioreq_evtchn; + struct rangeset *range[NR_IO_RANGE_TYPES]; + bool enabled; + uint8_t bufioreq_handling; +}; + static inline paddr_t ioreq_mmio_first_byte(const ioreq_t *p) { return unlikely(p->df) ? @@ -77,9 +111,9 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v); void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v); void hvm_destroy_all_ioreq_servers(struct domain *d); -struct hvm_ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p); -int hvm_send_ioreq(struct hvm_ioreq_server *s, ioreq_t *proto_p, +struct ioreq_server *hvm_select_ioreq_server(struct domain *d, + ioreq_t *p); +int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, bool buffered); unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); @@ -87,16 +121,16 @@ void hvm_ioreq_init(struct domain *d); bool arch_ioreq_complete_mmio(void); bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); -int arch_ioreq_server_map_pages(struct hvm_ioreq_server *s); -void arch_ioreq_server_unmap_pages(struct hvm_ioreq_server *s); -void arch_ioreq_server_enable(struct hvm_ioreq_server *s); -void arch_ioreq_server_disable(struct hvm_ioreq_server *s); -void arch_ioreq_server_destroy(struct hvm_ioreq_server *s); +int arch_ioreq_server_map_pages(struct ioreq_server *s); +void arch_ioreq_server_unmap_pages(struct ioreq_server *s); +void arch_ioreq_server_enable(struct ioreq_server *s); +void arch_ioreq_server_disable(struct ioreq_server *s); +void arch_ioreq_server_destroy(struct ioreq_server *s); int arch_ioreq_server_map_mem_type(struct domain *d, - struct hvm_ioreq_server *s, + struct ioreq_server *s, uint32_t flags); void arch_ioreq_server_map_mem_type_completed(struct domain *d, - struct hvm_ioreq_server *s, + struct ioreq_server *s, uint32_t flags); bool arch_ioreq_server_destroy_all(struct domain *d); bool arch_ioreq_server_get_type_addr(const struct domain *d, const ioreq_t *p, From patchwork Mon Jan 25 19:08:15 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044047 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id F40E4C433E0 for ; Mon, 25 Jan 2021 19:09:43 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 9F9912100A for ; Mon, 25 Jan 2021 19:09:43 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 9F9912100A Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74306.133597 (Exim 4.92) (envelope-from ) id 1l47F2-0000aE-O7; Mon, 25 Jan 2021 19:09:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74306.133597; Mon, 25 Jan 2021 19:09:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47F2-0000a6-IR; Mon, 25 Jan 2021 19:09:36 +0000 Received: by outflank-mailman (input) for mailman id 74306; Mon, 25 Jan 2021 19:09:35 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47F1-0008N7-GW for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:35 +0000 Received: from mail-wr1-x432.google.com (unknown [2a00:1450:4864:20::432]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 57e928a5-a02d-446b-840c-138bd6d88fb1; Mon, 25 Jan 2021 19:09:02 +0000 (UTC) Received: by mail-wr1-x432.google.com with SMTP id g10so14126606wrx.1 for ; Mon, 25 Jan 2021 11:09:02 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.00 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:00 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 57e928a5-a02d-446b-840c-138bd6d88fb1 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=WXYwiN3dYkxs2ItAbEK7clRgBID+kN/cztiUPQCB6UM=; b=Zzjx+qG0CtVKWbhpDzOa/ZbGehbgh0a5J3yewHX7TR06y/MA5aXWnRhxrpJEL378+R gxABx92HyCU1YGkZl9CxqK29y5aLpQOoTOh1hp6/njGwZNwQyS7ApHAUiU8RlpR5YSlc 9g6YICYeHmK1zMtDhhMprunhCBZ8wRnuZBIJSbKfr800lebxKGeUfP2en9Npw58e9SgA vsJmmW3/kyi0U0XnkvXL7XkFMHMaNEVYjTzIqPE+9TyuqLRnZ5PbjP5fYoZ7jkZpKCTy qwwchDQfAyT/0oHNwMbZs//sw4L6h+9Ate7uTTEoCW6B8mh7BHmT7A2mPBSkc6fTE3Sc jfZw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=WXYwiN3dYkxs2ItAbEK7clRgBID+kN/cztiUPQCB6UM=; b=NTCHJKJJbrvALhwQa0HciaYM0qCpKnt0bCoDpCrrxUe1MpAdWbydsNm1qfX+cfcy3n Lx61iBqWNIroh5jHe5GMdlRsDr7+iJft2awLEfsWF46ExbwJQEN4KUkVDz3w/bq6iNpB MfRdTCzJ2WIxMmyxmO8cN9TwRXgyRy08x1DO5Hz0RrfGILHok8W2qcM2PwsX6GeHCX9t 7IhXdXdG+cuDIhKTwXk9n14iaaN2s9l4TKonBzIxBJ9C6gJ3ZYwk1mCstNxF9lBcliVh 4mJVbcL5sEYoJSehqLfY3j+c3dAyk0ELExi9AxroT7YfDVk0yXrZMkaWb/n648GVhb+N kvhA== X-Gm-Message-State: AOAM531IzOD9d7xDIHDZFCKugLcypNuBluZFagCkeAdeY7h5Jw5hm+1b LwY1b6HpOfJ4gF5itAd4bRJi+1zRmhxM+w== X-Google-Smtp-Source: ABdhPJxMlQeyE6eQ2sYCm8DaJIzIl5PuEeAXb1GklBXXJsYAfCsff6PMXMzTZP16GF2R916XqXqQyg== X-Received: by 2002:a05:6000:1142:: with SMTP id d2mr2596225wrx.307.1611601741237; Mon, 25 Jan 2021 11:09:01 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Andrew Cooper , George Dunlap , Ian Jackson , Jan Beulich , Julien Grall , Stefano Stabellini , Wei Liu , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Julien Grall Subject: [PATCH V5 08/22] xen/ioreq: Move x86's ioreq_server to struct domain Date: Mon, 25 Jan 2021 21:08:15 +0200 Message-Id: <1611601709-28361-9-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko The IOREQ is a common feature now and this struct will be used on Arm as is. Move it to common struct domain. This also significantly reduces the layering violation in the common code (*arch.hvm* usage). We don't move ioreq_gfn since it is not used in the common code (the "legacy" mechanism is x86 specific). Signed-off-by: Oleksandr Tyshchenko Acked-by: Jan Beulich Reviewed-by: Julien Grall Reviewed-by: Paul Durrant Reviewed-by: Alex Bennée CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V1 -> V2: - new patch Changes V2 -> V3: - remove the mention of "ioreq_gfn" from patch subject/description - update patch according the "legacy interface" is x86 specific - drop hvm_params related changes in arch/x86/hvm/hvm.c - leave ioreq_gfn in hvm_domain Changes V3 -> V4: - rebase - drop the stale part of the comment above struct ioreq_server - add Jan's A-b Changes V4 -> V5: - add Julien's, Alex's and Paul's R-b --- --- xen/common/ioreq.c | 60 ++++++++++++++++++++-------------------- xen/include/asm-x86/hvm/domain.h | 8 ------ xen/include/xen/sched.h | 10 +++++++ 3 files changed, 40 insertions(+), 38 deletions(-) diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 7320f23..4cb26e6 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -38,13 +38,13 @@ static void set_ioreq_server(struct domain *d, unsigned int id, struct ioreq_server *s) { ASSERT(id < MAX_NR_IOREQ_SERVERS); - ASSERT(!s || !d->arch.hvm.ioreq_server.server[id]); + ASSERT(!s || !d->ioreq_server.server[id]); - d->arch.hvm.ioreq_server.server[id] = s; + d->ioreq_server.server[id] = s; } #define GET_IOREQ_SERVER(d, id) \ - (d)->arch.hvm.ioreq_server.server[id] + (d)->ioreq_server.server[id] static struct ioreq_server *get_ioreq_server(const struct domain *d, unsigned int id) @@ -285,7 +285,7 @@ bool is_ioreq_server_page(struct domain *d, const struct page_info *page) unsigned int id; bool found = false; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); FOR_EACH_IOREQ_SERVER(d, id, s) { @@ -296,7 +296,7 @@ bool is_ioreq_server_page(struct domain *d, const struct page_info *page) } } - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return found; } @@ -606,7 +606,7 @@ int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, return -ENOMEM; domain_pause(d); - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); for ( i = 0; i < MAX_NR_IOREQ_SERVERS; i++ ) { @@ -634,13 +634,13 @@ int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, if ( id ) *id = i; - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); domain_unpause(d); return 0; fail: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); domain_unpause(d); xfree(s); @@ -652,7 +652,7 @@ int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) struct ioreq_server *s; int rc; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -684,7 +684,7 @@ int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) rc = 0; out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -697,7 +697,7 @@ int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, struct ioreq_server *s; int rc; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -731,7 +731,7 @@ int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, rc = 0; out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -744,7 +744,7 @@ int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, ASSERT(is_hvm_domain(d)); - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -782,7 +782,7 @@ int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, } out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -798,7 +798,7 @@ int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, if ( start > end ) return -EINVAL; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -834,7 +834,7 @@ int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, rc = rangeset_add_range(r, start, end); out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -850,7 +850,7 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, if ( start > end ) return -EINVAL; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -886,7 +886,7 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, rc = rangeset_remove_range(r, start, end); out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -911,7 +911,7 @@ int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, if ( flags & ~XEN_DMOP_IOREQ_MEM_ACCESS_WRITE ) return -EINVAL; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -926,7 +926,7 @@ int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, rc = arch_ioreq_server_map_mem_type(d, s, flags); out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); if ( rc == 0 ) arch_ioreq_server_map_mem_type_completed(d, s, flags); @@ -940,7 +940,7 @@ int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, struct ioreq_server *s; int rc; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); s = get_ioreq_server(d, id); @@ -964,7 +964,7 @@ int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, rc = 0; out: - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -974,7 +974,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) unsigned int id; int rc; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); FOR_EACH_IOREQ_SERVER(d, id, s) { @@ -983,7 +983,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) goto fail; } - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return 0; @@ -998,7 +998,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) hvm_ioreq_server_remove_vcpu(s, v); } - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); return rc; } @@ -1008,12 +1008,12 @@ void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) struct ioreq_server *s; unsigned int id; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); FOR_EACH_IOREQ_SERVER(d, id, s) hvm_ioreq_server_remove_vcpu(s, v); - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); } void hvm_destroy_all_ioreq_servers(struct domain *d) @@ -1024,7 +1024,7 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) if ( !arch_ioreq_server_destroy_all(d) ) return; - spin_lock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_lock_recursive(&d->ioreq_server.lock); /* No need to domain_pause() as the domain is being torn down */ @@ -1042,7 +1042,7 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) xfree(s); } - spin_unlock_recursive(&d->arch.hvm.ioreq_server.lock); + spin_unlock_recursive(&d->ioreq_server.lock); } struct ioreq_server *hvm_select_ioreq_server(struct domain *d, @@ -1274,7 +1274,7 @@ unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) void hvm_ioreq_init(struct domain *d) { - spin_lock_init(&d->arch.hvm.ioreq_server.lock); + spin_lock_init(&d->ioreq_server.lock); arch_ioreq_domain_init(d); } diff --git a/xen/include/asm-x86/hvm/domain.h b/xen/include/asm-x86/hvm/domain.h index 3b36c2f..b8be1ad 100644 --- a/xen/include/asm-x86/hvm/domain.h +++ b/xen/include/asm-x86/hvm/domain.h @@ -63,8 +63,6 @@ struct hvm_pi_ops { void (*vcpu_block)(struct vcpu *); }; -#define MAX_NR_IOREQ_SERVERS 8 - struct hvm_domain { /* Guest page range used for non-default ioreq servers */ struct { @@ -73,12 +71,6 @@ struct hvm_domain { unsigned long legacy_mask; /* indexed by HVM param number */ } ioreq_gfn; - /* Lock protects all other values in the sub-struct and the default */ - struct { - spinlock_t lock; - struct ioreq_server *server[MAX_NR_IOREQ_SERVERS]; - } ioreq_server; - /* Cached CF8 for guest PCI config cycles */ uint32_t pci_cf8; diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index da19f4e..f437ee3 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -318,6 +318,8 @@ struct sched_unit { struct evtchn_port_ops; +#define MAX_NR_IOREQ_SERVERS 8 + struct domain { domid_t domain_id; @@ -534,6 +536,14 @@ struct domain unsigned int val; struct vcpu *vcpu; } teardown; + +#ifdef CONFIG_IOREQ_SERVER + /* Lock protects all other values in the sub-struct */ + struct { + spinlock_t lock; + struct ioreq_server *server[MAX_NR_IOREQ_SERVERS]; + } ioreq_server; +#endif }; static inline struct page_list_head *page_to_list( From patchwork Mon Jan 25 19:08:16 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044089 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 58CA9C433E9 for ; Mon, 25 Jan 2021 19:20:11 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id ED57C21D94 for ; Mon, 25 Jan 2021 19:20:10 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org ED57C21D94 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74350.133651 (Exim 4.92) (envelope-from ) id 1l47P6-00027g-4V; Mon, 25 Jan 2021 19:20:00 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74350.133651; Mon, 25 Jan 2021 19:20:00 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P5-00027A-S9; Mon, 25 Jan 2021 19:19:59 +0000 Received: by outflank-mailman (input) for mailman id 74350; Mon, 25 Jan 2021 19:19:58 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47FB-0008N7-H1 for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:45 +0000 Received: from mail-wr1-x42f.google.com (unknown [2a00:1450:4864:20::42f]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id cd2505de-ed76-4c20-b73f-826dfae0fcf2; Mon, 25 Jan 2021 19:09:03 +0000 (UTC) Received: by mail-wr1-x42f.google.com with SMTP id q7so14089321wre.13 for ; Mon, 25 Jan 2021 11:09:03 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.01 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:01 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: cd2505de-ed76-4c20-b73f-826dfae0fcf2 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=gfzEe12oOQQBTndDYAH9xFkPRyp+VWBaMh+oyxymM3E=; b=rUhJC4FwzZsW35dnwXht4R1/QOBAKbF3wN72NkbcNt7GTn5i0ygeHOoguVTyoq3Q9F ToqeuSINhN8HMj8UpSg16lv+ZlHVw24NifG9pVvbwVGSM/GFyVgpZKjsOzpzFFNl3V4k HRjGdvTuBuLWEY7SEa0fETsnXCV169aJHOXeOUmK+3huZO3wOA7uUmqMehFkyfe75YdZ wJjjIUl+3ss5kCdUopbiuYrnlvl1IhEKUc94nskTh+fH553O4l+N2cDwcDU7CsiEvalF oUDn14jHhNJLtAlIgYuIJcI1wKcfNfTteQmZDpSgubTapLHOGkSdy454qz3+c/dXFkYP Ag+A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=gfzEe12oOQQBTndDYAH9xFkPRyp+VWBaMh+oyxymM3E=; b=O9LMQfwQ6JJZsrnQirAlEic8LSwLTV2mU0gcCCnh0M9s82J/zKg2Z3LjHwJAV21Tcr qafK+5aEzQYIH3gJyFhQvXlwJN+QwtZ29qwOlyE8KZfdV1FQL4Tg3BdPdMNITp57gVLp 4vh1DGR9SjX0AvyQIi365IjOYXI6R7CWdjsa4dUb254oH4PMcluRTSWm0DpYvZKqsYeL y9qBI8JuPsJR+c4jLjAQWJxRHonGD5mh9O07CnzGxW2xo3SH1FLxuMQmFY/fN9wHRhoL F2PKog/e500OmMuM7CNKG2mVkBGXbKPS9TO9qqHJFbm05oulkQb93KR0cPx/C4Jw02hc zMBA== X-Gm-Message-State: AOAM533a5vXWz/3dFLhro8eR3wbGFfYzXparJPqAOL87dde/V1IpRHbO GqD0hvKzHFkcxDwHBBn5o/Q4BUWddVDN8Q== X-Google-Smtp-Source: ABdhPJwa03FnqY1HkcHrA6ZQVrPqf+MqnBXBsRDtaVhpB1KjWEYjBKOl2dXeMbTpz0lIm1IsFzGfiw== X-Received: by 2002:adf:e448:: with SMTP id t8mr2606061wrm.288.1611601742356; Mon, 25 Jan 2021 11:09:02 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Julien Grall , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Ian Jackson , Julien Grall , Stefano Stabellini , Paul Durrant , Daniel De Graaf , Oleksandr Tyshchenko Subject: [PATCH V5 09/22] xen/ioreq: Make x86's IOREQ related dm-op handling common Date: Mon, 25 Jan 2021 21:08:16 +0200 Message-Id: <1611601709-28361-10-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Julien Grall As a lot of x86 code can be re-used on Arm later on, this patch moves the IOREQ related dm-op handling to the common code. The idea is to have the top level dm-op handling arch-specific and call into ioreq_server_dm_op() for otherwise unhandled ops. Pros: - More natural than doing it other way around (top level dm-op handling common). - Leave compat_dm_op() in x86 code. Cons: - Code duplication. Both arches have to duplicate dm_op(), etc. Make the corresponding functions static and rename them according to the new naming scheme (including dropping the "hvm" prefixes). Introduce common dm.c file as a resting place for the do_dm_op() (which is identical for both Arm and x86) to minimize code duplication. The common DM feature is supposed to be built with IOREQ_SERVER option enabled (as well as the IOREQ feature), which is selected for x86's config HVM for now. Also update XSM code a bit to let dm-op be used on Arm. This support is going to be used on Arm to be able run device emulator outside of Xen hypervisor. Signed-off-by: Julien Grall Signed-off-by: Oleksandr Tyshchenko Acked-by: Jan Beulich [On Arm only] Tested-by: Wei Chen Reviewed-by: Paul Durrant --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - update XSM, related changes were pulled from: [RFC PATCH V1 04/12] xen/arm: Introduce arch specific bits for IOREQ/DM features Changes V1 -> V2: - update the author of a patch - update patch description - introduce xen/dm.h and move definitions here Changes V2 -> V3: - no changes Changes V3 -> V4: - rework to have the top level dm-op handling arch-specific - update patch subject/description, was "xen/dm: Make x86's DM feature common" - make a few functions static in common ioreq.c Changes V4 -> V5: - update patch description - add Jan's A-b - drop the 'hvm_' prefixes of touched functions and rename them instead of doing that in patch #12 - add common dm.c to keep do_dm_op(), make dm_op() global --- --- xen/arch/x86/hvm/dm.c | 132 ++-------------------------------------------- xen/common/Makefile | 1 + xen/common/dm.c | 55 +++++++++++++++++++ xen/common/ioreq.c | 137 ++++++++++++++++++++++++++++++++++++++++++------ xen/include/xen/dm.h | 41 +++++++++++++++ xen/include/xen/ioreq.h | 21 ++------ xen/include/xsm/dummy.h | 4 +- xen/include/xsm/xsm.h | 6 +-- xen/xsm/dummy.c | 2 +- xen/xsm/flask/hooks.c | 5 +- 10 files changed, 235 insertions(+), 169 deletions(-) create mode 100644 xen/common/dm.c create mode 100644 xen/include/xen/dm.h diff --git a/xen/arch/x86/hvm/dm.c b/xen/arch/x86/hvm/dm.c index d3e2a9e..5bc172a 100644 --- a/xen/arch/x86/hvm/dm.c +++ b/xen/arch/x86/hvm/dm.c @@ -16,6 +16,7 @@ #include #include +#include #include #include #include @@ -29,13 +30,6 @@ #include -struct dmop_args { - domid_t domid; - unsigned int nr_bufs; - /* Reserve enough buf elements for all current hypercalls. */ - struct xen_dm_op_buf buf[2]; -}; - static bool _raw_copy_from_guest_buf_offset(void *dst, const struct dmop_args *args, unsigned int buf_idx, @@ -338,7 +332,7 @@ static int inject_event(struct domain *d, return 0; } -static int dm_op(const struct dmop_args *op_args) +int dm_op(const struct dmop_args *op_args) { struct domain *d; struct xen_dm_op op; @@ -408,71 +402,6 @@ static int dm_op(const struct dmop_args *op_args) switch ( op.op ) { - case XEN_DMOP_create_ioreq_server: - { - struct xen_dm_op_create_ioreq_server *data = - &op.u.create_ioreq_server; - - const_op = false; - - rc = -EINVAL; - if ( data->pad[0] || data->pad[1] || data->pad[2] ) - break; - - rc = hvm_create_ioreq_server(d, data->handle_bufioreq, - &data->id); - break; - } - - case XEN_DMOP_get_ioreq_server_info: - { - struct xen_dm_op_get_ioreq_server_info *data = - &op.u.get_ioreq_server_info; - const uint16_t valid_flags = XEN_DMOP_no_gfns; - - const_op = false; - - rc = -EINVAL; - if ( data->flags & ~valid_flags ) - break; - - rc = hvm_get_ioreq_server_info(d, data->id, - (data->flags & XEN_DMOP_no_gfns) ? - NULL : &data->ioreq_gfn, - (data->flags & XEN_DMOP_no_gfns) ? - NULL : &data->bufioreq_gfn, - &data->bufioreq_port); - break; - } - - case XEN_DMOP_map_io_range_to_ioreq_server: - { - const struct xen_dm_op_ioreq_server_range *data = - &op.u.map_io_range_to_ioreq_server; - - rc = -EINVAL; - if ( data->pad ) - break; - - rc = hvm_map_io_range_to_ioreq_server(d, data->id, data->type, - data->start, data->end); - break; - } - - case XEN_DMOP_unmap_io_range_from_ioreq_server: - { - const struct xen_dm_op_ioreq_server_range *data = - &op.u.unmap_io_range_from_ioreq_server; - - rc = -EINVAL; - if ( data->pad ) - break; - - rc = hvm_unmap_io_range_from_ioreq_server(d, data->id, data->type, - data->start, data->end); - break; - } - case XEN_DMOP_map_mem_type_to_ioreq_server: { struct xen_dm_op_map_mem_type_to_ioreq_server *data = @@ -486,8 +415,8 @@ static int dm_op(const struct dmop_args *op_args) break; if ( first_gfn == 0 ) - rc = hvm_map_mem_type_to_ioreq_server(d, data->id, - data->type, data->flags); + rc = ioreq_server_map_mem_type(d, data->id, + data->type, data->flags); else rc = 0; @@ -523,32 +452,6 @@ static int dm_op(const struct dmop_args *op_args) break; } - case XEN_DMOP_set_ioreq_server_state: - { - const struct xen_dm_op_set_ioreq_server_state *data = - &op.u.set_ioreq_server_state; - - rc = -EINVAL; - if ( data->pad ) - break; - - rc = hvm_set_ioreq_server_state(d, data->id, !!data->enabled); - break; - } - - case XEN_DMOP_destroy_ioreq_server: - { - const struct xen_dm_op_destroy_ioreq_server *data = - &op.u.destroy_ioreq_server; - - rc = -EINVAL; - if ( data->pad ) - break; - - rc = hvm_destroy_ioreq_server(d, data->id); - break; - } - case XEN_DMOP_track_dirty_vram: { const struct xen_dm_op_track_dirty_vram *data = @@ -703,7 +606,7 @@ static int dm_op(const struct dmop_args *op_args) } default: - rc = -EOPNOTSUPP; + rc = ioreq_server_dm_op(&op, d, &const_op); break; } @@ -776,31 +679,6 @@ int compat_dm_op(domid_t domid, return rc; } -long do_dm_op(domid_t domid, - unsigned int nr_bufs, - XEN_GUEST_HANDLE_PARAM(xen_dm_op_buf_t) bufs) -{ - struct dmop_args args; - int rc; - - if ( nr_bufs > ARRAY_SIZE(args.buf) ) - return -E2BIG; - - args.domid = domid; - args.nr_bufs = array_index_nospec(nr_bufs, ARRAY_SIZE(args.buf) + 1); - - if ( copy_from_guest_offset(&args.buf[0], bufs, 0, args.nr_bufs) ) - return -EFAULT; - - rc = dm_op(&args); - - if ( rc == -ERESTART ) - rc = hypercall_create_continuation(__HYPERVISOR_dm_op, "iih", - domid, nr_bufs, bufs); - - return rc; -} - /* * Local variables: * mode: C diff --git a/xen/common/Makefile b/xen/common/Makefile index 2e3c234..71c1d46 100644 --- a/xen/common/Makefile +++ b/xen/common/Makefile @@ -5,6 +5,7 @@ obj-$(CONFIG_CORE_PARKING) += core_parking.o obj-y += cpu.o obj-$(CONFIG_DEBUG_TRACE) += debugtrace.o obj-$(CONFIG_HAS_DEVICE_TREE) += device_tree.o +obj-$(CONFIG_IOREQ_SERVER) += dm.o obj-y += domain.o obj-y += event_2l.o obj-y += event_channel.o diff --git a/xen/common/dm.c b/xen/common/dm.c new file mode 100644 index 0000000..2d1d98c --- /dev/null +++ b/xen/common/dm.c @@ -0,0 +1,55 @@ +/* + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#include +#include +#include +#include + +long do_dm_op(domid_t domid, + unsigned int nr_bufs, + XEN_GUEST_HANDLE_PARAM(xen_dm_op_buf_t) bufs) +{ + struct dmop_args args; + int rc; + + if ( nr_bufs > ARRAY_SIZE(args.buf) ) + return -E2BIG; + + args.domid = domid; + args.nr_bufs = array_index_nospec(nr_bufs, ARRAY_SIZE(args.buf) + 1); + + if ( copy_from_guest_offset(&args.buf[0], bufs, 0, args.nr_bufs) ) + return -EFAULT; + + rc = dm_op(&args); + + if ( rc == -ERESTART ) + rc = hypercall_create_continuation(__HYPERVISOR_dm_op, "iih", + domid, nr_bufs, bufs); + + return rc; +} + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 4cb26e6..84bce36 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -223,7 +223,7 @@ static int hvm_alloc_ioreq_mfn(struct ioreq_server *s, bool buf) { /* * If a guest frame has already been mapped (which may happen - * on demand if hvm_get_ioreq_server_info() is called), then + * on demand if ioreq_server_get_info() is called), then * allocating a page is not permitted. */ if ( !gfn_eq(iorp->gfn, INVALID_GFN) ) @@ -591,8 +591,8 @@ static void hvm_ioreq_server_deinit(struct ioreq_server *s) put_domain(s->emulator); } -int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, - ioservid_t *id) +static int ioreq_server_create(struct domain *d, int bufioreq_handling, + ioservid_t *id) { struct ioreq_server *s; unsigned int i; @@ -647,7 +647,7 @@ int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, return rc; } -int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) +static int ioreq_server_destroy(struct domain *d, ioservid_t id) { struct ioreq_server *s; int rc; @@ -689,10 +689,10 @@ int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id) return rc; } -int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, - unsigned long *ioreq_gfn, - unsigned long *bufioreq_gfn, - evtchn_port_t *bufioreq_port) +static int ioreq_server_get_info(struct domain *d, ioservid_t id, + unsigned long *ioreq_gfn, + unsigned long *bufioreq_gfn, + evtchn_port_t *bufioreq_port) { struct ioreq_server *s; int rc; @@ -787,7 +787,7 @@ int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, return rc; } -int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, +static int ioreq_server_map_io_range(struct domain *d, ioservid_t id, uint32_t type, uint64_t start, uint64_t end) { @@ -839,9 +839,9 @@ int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, return rc; } -int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end) +static int ioreq_server_unmap_io_range(struct domain *d, ioservid_t id, + uint32_t type, uint64_t start, + uint64_t end) { struct ioreq_server *s; struct rangeset *r; @@ -899,8 +899,8 @@ int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, * Support for the emulation of read operations can be added when an ioreq * server has such requirement in the future. */ -int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint32_t flags) +int ioreq_server_map_mem_type(struct domain *d, ioservid_t id, + uint32_t type, uint32_t flags) { struct ioreq_server *s; int rc; @@ -934,8 +934,8 @@ int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, return rc; } -int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, - bool enabled) +static int ioreq_server_set_state(struct domain *d, ioservid_t id, + bool enabled) { struct ioreq_server *s; int rc; @@ -1279,6 +1279,111 @@ void hvm_ioreq_init(struct domain *d) arch_ioreq_domain_init(d); } +int ioreq_server_dm_op(struct xen_dm_op *op, struct domain *d, bool *const_op) +{ + long rc; + + switch ( op->op ) + { + case XEN_DMOP_create_ioreq_server: + { + struct xen_dm_op_create_ioreq_server *data = + &op->u.create_ioreq_server; + + *const_op = false; + + rc = -EINVAL; + if ( data->pad[0] || data->pad[1] || data->pad[2] ) + break; + + rc = ioreq_server_create(d, data->handle_bufioreq, + &data->id); + break; + } + + case XEN_DMOP_get_ioreq_server_info: + { + struct xen_dm_op_get_ioreq_server_info *data = + &op->u.get_ioreq_server_info; + const uint16_t valid_flags = XEN_DMOP_no_gfns; + + *const_op = false; + + rc = -EINVAL; + if ( data->flags & ~valid_flags ) + break; + + rc = ioreq_server_get_info(d, data->id, + (data->flags & XEN_DMOP_no_gfns) ? + NULL : (unsigned long *)&data->ioreq_gfn, + (data->flags & XEN_DMOP_no_gfns) ? + NULL : (unsigned long *)&data->bufioreq_gfn, + &data->bufioreq_port); + break; + } + + case XEN_DMOP_map_io_range_to_ioreq_server: + { + const struct xen_dm_op_ioreq_server_range *data = + &op->u.map_io_range_to_ioreq_server; + + rc = -EINVAL; + if ( data->pad ) + break; + + rc = ioreq_server_map_io_range(d, data->id, data->type, + data->start, data->end); + break; + } + + case XEN_DMOP_unmap_io_range_from_ioreq_server: + { + const struct xen_dm_op_ioreq_server_range *data = + &op->u.unmap_io_range_from_ioreq_server; + + rc = -EINVAL; + if ( data->pad ) + break; + + rc = ioreq_server_unmap_io_range(d, data->id, data->type, + data->start, data->end); + break; + } + + case XEN_DMOP_set_ioreq_server_state: + { + const struct xen_dm_op_set_ioreq_server_state *data = + &op->u.set_ioreq_server_state; + + rc = -EINVAL; + if ( data->pad ) + break; + + rc = ioreq_server_set_state(d, data->id, !!data->enabled); + break; + } + + case XEN_DMOP_destroy_ioreq_server: + { + const struct xen_dm_op_destroy_ioreq_server *data = + &op->u.destroy_ioreq_server; + + rc = -EINVAL; + if ( data->pad ) + break; + + rc = ioreq_server_destroy(d, data->id); + break; + } + + default: + rc = -EOPNOTSUPP; + break; + } + + return rc; +} + /* * Local variables: * mode: C diff --git a/xen/include/xen/dm.h b/xen/include/xen/dm.h new file mode 100644 index 0000000..4bd7e57 --- /dev/null +++ b/xen/include/xen/dm.h @@ -0,0 +1,41 @@ +/* + * Copyright (c) 2016 Citrix Systems Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#ifndef __XEN_DM_H__ +#define __XEN_DM_H__ + +#include + +struct dmop_args { + domid_t domid; + unsigned int nr_bufs; + /* Reserve enough buf elements for all current hypercalls. */ + struct xen_dm_op_buf buf[2]; +}; + +int dm_op(const struct dmop_args *op_args); + +#endif /* __XEN_DM_H__ */ + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 5a6c11d..60e864d 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -87,25 +87,10 @@ bool hvm_io_pending(struct vcpu *v); bool handle_hvm_io_completion(struct vcpu *v); bool is_ioreq_server_page(struct domain *d, const struct page_info *page); -int hvm_create_ioreq_server(struct domain *d, int bufioreq_handling, - ioservid_t *id); -int hvm_destroy_ioreq_server(struct domain *d, ioservid_t id); -int hvm_get_ioreq_server_info(struct domain *d, ioservid_t id, - unsigned long *ioreq_gfn, - unsigned long *bufioreq_gfn, - evtchn_port_t *bufioreq_port); int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, unsigned long idx, mfn_t *mfn); -int hvm_map_io_range_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end); -int hvm_unmap_io_range_from_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint64_t start, - uint64_t end); -int hvm_map_mem_type_to_ioreq_server(struct domain *d, ioservid_t id, - uint32_t type, uint32_t flags); -int hvm_set_ioreq_server_state(struct domain *d, ioservid_t id, - bool enabled); +int ioreq_server_map_mem_type(struct domain *d, ioservid_t id, + uint32_t type, uint32_t flags); int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v); void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v); @@ -119,6 +104,8 @@ unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); void hvm_ioreq_init(struct domain *d); +int ioreq_server_dm_op(struct xen_dm_op *op, struct domain *d, bool *const_op); + bool arch_ioreq_complete_mmio(void); bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); int arch_ioreq_server_map_pages(struct ioreq_server *s); diff --git a/xen/include/xsm/dummy.h b/xen/include/xsm/dummy.h index fa40e88..10739e7 100644 --- a/xen/include/xsm/dummy.h +++ b/xen/include/xsm/dummy.h @@ -707,14 +707,14 @@ static XSM_INLINE int xsm_pmu_op (XSM_DEFAULT_ARG struct domain *d, unsigned int } } +#endif /* CONFIG_X86 */ + static XSM_INLINE int xsm_dm_op(XSM_DEFAULT_ARG struct domain *d) { XSM_ASSERT_ACTION(XSM_DM_PRIV); return xsm_default_action(action, current->domain, d); } -#endif /* CONFIG_X86 */ - #ifdef CONFIG_ARGO static XSM_INLINE int xsm_argo_enable(const struct domain *d) { diff --git a/xen/include/xsm/xsm.h b/xen/include/xsm/xsm.h index 7bd03d8..91ecff4 100644 --- a/xen/include/xsm/xsm.h +++ b/xen/include/xsm/xsm.h @@ -176,8 +176,8 @@ struct xsm_operations { int (*ioport_permission) (struct domain *d, uint32_t s, uint32_t e, uint8_t allow); int (*ioport_mapping) (struct domain *d, uint32_t s, uint32_t e, uint8_t allow); int (*pmu_op) (struct domain *d, unsigned int op); - int (*dm_op) (struct domain *d); #endif + int (*dm_op) (struct domain *d); int (*xen_version) (uint32_t cmd); int (*domain_resource_map) (struct domain *d); #ifdef CONFIG_ARGO @@ -682,13 +682,13 @@ static inline int xsm_pmu_op (xsm_default_t def, struct domain *d, unsigned int return xsm_ops->pmu_op(d, op); } +#endif /* CONFIG_X86 */ + static inline int xsm_dm_op(xsm_default_t def, struct domain *d) { return xsm_ops->dm_op(d); } -#endif /* CONFIG_X86 */ - static inline int xsm_xen_version (xsm_default_t def, uint32_t op) { return xsm_ops->xen_version(op); diff --git a/xen/xsm/dummy.c b/xen/xsm/dummy.c index 9e09512..8bdffe7 100644 --- a/xen/xsm/dummy.c +++ b/xen/xsm/dummy.c @@ -147,8 +147,8 @@ void __init xsm_fixup_ops (struct xsm_operations *ops) set_to_dummy_if_null(ops, ioport_permission); set_to_dummy_if_null(ops, ioport_mapping); set_to_dummy_if_null(ops, pmu_op); - set_to_dummy_if_null(ops, dm_op); #endif + set_to_dummy_if_null(ops, dm_op); set_to_dummy_if_null(ops, xen_version); set_to_dummy_if_null(ops, domain_resource_map); #ifdef CONFIG_ARGO diff --git a/xen/xsm/flask/hooks.c b/xen/xsm/flask/hooks.c index 19b0d9e..11784d7 100644 --- a/xen/xsm/flask/hooks.c +++ b/xen/xsm/flask/hooks.c @@ -1656,14 +1656,13 @@ static int flask_pmu_op (struct domain *d, unsigned int op) return -EPERM; } } +#endif /* CONFIG_X86 */ static int flask_dm_op(struct domain *d) { return current_has_perm(d, SECCLASS_HVM, HVM__DM); } -#endif /* CONFIG_X86 */ - static int flask_xen_version (uint32_t op) { u32 dsid = domain_sid(current->domain); @@ -1865,8 +1864,8 @@ static struct xsm_operations flask_ops = { .ioport_permission = flask_ioport_permission, .ioport_mapping = flask_ioport_mapping, .pmu_op = flask_pmu_op, - .dm_op = flask_dm_op, #endif + .dm_op = flask_dm_op, .xen_version = flask_xen_version, .domain_resource_map = flask_domain_resource_map, #ifdef CONFIG_ARGO From patchwork Mon Jan 25 19:08:17 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044093 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 8F45AC433E6 for ; Mon, 25 Jan 2021 19:20:13 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 3A4CA21744 for ; Mon, 25 Jan 2021 19:20:13 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 3A4CA21744 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74355.133701 (Exim 4.92) (envelope-from ) id 1l47P9-0002IQ-OY; Mon, 25 Jan 2021 19:20:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74355.133701; Mon, 25 Jan 2021 19:20:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P9-0002Hu-5u; Mon, 25 Jan 2021 19:20:03 +0000 Received: by outflank-mailman (input) for mailman id 74355; Mon, 25 Jan 2021 19:20:00 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47FL-0008N7-HR for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:55 +0000 Received: from mail-wr1-x42a.google.com (unknown [2a00:1450:4864:20::42a]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 5a44a0f1-3cec-476d-9480-653689a238c9; Mon, 25 Jan 2021 19:09:04 +0000 (UTC) Received: by mail-wr1-x42a.google.com with SMTP id c12so14120629wrc.7 for ; Mon, 25 Jan 2021 11:09:04 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.02 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:03 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 5a44a0f1-3cec-476d-9480-653689a238c9 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=WLZbCR4gkU6sLTn4NNkQ608UvyBwuWACZ5paSpJruFE=; b=rfGYH3GUoppFqAMP68cfP3W83RNpGUmiTOw+gvmcrMr8YOPIKWaNDMbXYCf6AgjopM +KuleFkNB0LHkKatiabCHd98TkxEOVqhN59H76gX6mPAD7HpDY4YR1aV8X1erEiMw3MC qQusvAoIx+QfFFHxpwrSKr7wzGNUugCXn36vH5Gi4Qo8oEsV87Gn5Wj/DUMQcm2Tt4Jl 4AczD8h6qJet9rvEy4V0Fy4RS/NinbfJkI3UTbIp0mA1NP2Ejjt8LaVcqiSNGByjmbyu QDcLGKOubFemMF4FzWbJmEKp5sfBi6jnOcbN1onYLtZIcqRL3isNTh1nfgVVko2FUH3R V0yg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=WLZbCR4gkU6sLTn4NNkQ608UvyBwuWACZ5paSpJruFE=; b=ervnsG+SJbh+PBfIG3tikinLXRBmTcwFQ3MusYdnidnLiw9BsHNuXKcyo1g7EbNMg4 MXxNUpvCSeSAfhRiF74ExSUeKErLvf3eLwiwCMC/0Yvx0ocgXnUX/pTGbsL0/Dme/55D XrqGqLwdWZp/b0gwHmRyf2JelXXWdPKJSZowTqr8xR8Du+blROr643YdPLf5llUerUbF +rg3RIjQhJWkLuL4TEcW9O8bftlPwSi2D1ZL2ym3aVPcODLQXOP4qi2CCzPXzW3Q5doK /fXLCSHXQF54q3btqheurJD0QGMIy+IzuLup1Z7ieJnrmpnGM48oqSphNvakFy8LRllm Oulw== X-Gm-Message-State: AOAM531zsgJB5yzYT1N1ys/yu4b2UXNPQTKPhCWWB7ejZiVA0fFrtNQR QKZQcjWUh2fMzG5LNdULOWZW+lBNdPcFRQ== X-Google-Smtp-Source: ABdhPJyOXKtUXpygg+Tv1S8trRlGXIhMh8yf/dp7Zhmbl7Xtx9KZXGRlJz5gUffRBeaMKU4ri/QXWw== X-Received: by 2002:a5d:5304:: with SMTP id e4mr2530906wrv.167.1611601743586; Mon, 25 Jan 2021 11:09:03 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Ian Jackson , Julien Grall , Stefano Stabellini , Jun Nakajima , Kevin Tian , Julien Grall Subject: [PATCH V5 10/22] xen/ioreq: Move x86's io_completion/io_req fields to struct vcpu Date: Mon, 25 Jan 2021 21:08:17 +0200 Message-Id: <1611601709-28361-11-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko The IOREQ is a common feature now and these fields will be used on Arm as is. Move them to common struct vcpu as a part of new struct vcpu_io and drop duplicating "io" prefixes. Also move enum hvm_io_completion to xen/sched.h and remove "hvm" prefixes. This patch completely removes layering violation in the common code. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Julien Grall Reviewed-by: Paul Durrant Acked-by: Jan Beulich CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V1 -> V2: - new patch Changes V2 -> V3: - update patch according the "legacy interface" is x86 specific - update patch description - drop the "io" prefixes from the field names - wrap IO_realmode_completion Changes V3 -> V4: - rename all hvm_vcpu_io locals to "hvio" - rename according to the new renaming scheme IO_ -> VIO_ (io_ -> vio_) - drop "io" prefix from io_completion locals Changes V4 -> V5: - add Julien's and Paul's R-b, Jan's A-b --- --- xen/arch/x86/hvm/emulate.c | 210 +++++++++++++++++++------------------- xen/arch/x86/hvm/hvm.c | 2 +- xen/arch/x86/hvm/io.c | 32 +++--- xen/arch/x86/hvm/ioreq.c | 6 +- xen/arch/x86/hvm/svm/nestedsvm.c | 2 +- xen/arch/x86/hvm/vmx/realmode.c | 8 +- xen/common/ioreq.c | 26 ++--- xen/include/asm-x86/hvm/emulate.h | 2 +- xen/include/asm-x86/hvm/vcpu.h | 11 -- xen/include/xen/ioreq.h | 2 +- xen/include/xen/sched.h | 19 ++++ 11 files changed, 164 insertions(+), 156 deletions(-) diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c index 4d62199..21051ce 100644 --- a/xen/arch/x86/hvm/emulate.c +++ b/xen/arch/x86/hvm/emulate.c @@ -140,15 +140,15 @@ static const struct hvm_io_handler ioreq_server_handler = { */ void hvmemul_cancel(struct vcpu *v) { - struct hvm_vcpu_io *vio = &v->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &v->arch.hvm.hvm_io; - vio->io_req.state = STATE_IOREQ_NONE; - vio->io_completion = HVMIO_no_completion; - vio->mmio_cache_count = 0; - vio->mmio_insn_bytes = 0; - vio->mmio_access = (struct npfec){}; - vio->mmio_retry = false; - vio->g2m_ioport = NULL; + v->io.req.state = STATE_IOREQ_NONE; + v->io.completion = VIO_no_completion; + hvio->mmio_cache_count = 0; + hvio->mmio_insn_bytes = 0; + hvio->mmio_access = (struct npfec){}; + hvio->mmio_retry = false; + hvio->g2m_ioport = NULL; hvmemul_cache_disable(v); } @@ -159,7 +159,7 @@ static int hvmemul_do_io( { struct vcpu *curr = current; struct domain *currd = curr->domain; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct vcpu_io *vio = &curr->io; ioreq_t p = { .type = is_mmio ? IOREQ_TYPE_COPY : IOREQ_TYPE_PIO, .addr = addr, @@ -184,13 +184,13 @@ static int hvmemul_do_io( return X86EMUL_UNHANDLEABLE; } - switch ( vio->io_req.state ) + switch ( vio->req.state ) { case STATE_IOREQ_NONE: break; case STATE_IORESP_READY: - vio->io_req.state = STATE_IOREQ_NONE; - p = vio->io_req; + vio->req.state = STATE_IOREQ_NONE; + p = vio->req; /* Verify the emulation request has been correctly re-issued */ if ( (p.type != (is_mmio ? IOREQ_TYPE_COPY : IOREQ_TYPE_PIO)) || @@ -238,7 +238,7 @@ static int hvmemul_do_io( } ASSERT(p.count); - vio->io_req = p; + vio->req = p; rc = hvm_io_intercept(&p); @@ -247,12 +247,12 @@ static int hvmemul_do_io( * our callers and mirror this into latched state. */ ASSERT(p.count <= *reps); - *reps = vio->io_req.count = p.count; + *reps = vio->req.count = p.count; switch ( rc ) { case X86EMUL_OKAY: - vio->io_req.state = STATE_IOREQ_NONE; + vio->req.state = STATE_IOREQ_NONE; break; case X86EMUL_UNHANDLEABLE: { @@ -305,7 +305,7 @@ static int hvmemul_do_io( if ( s == NULL ) { rc = X86EMUL_RETRY; - vio->io_req.state = STATE_IOREQ_NONE; + vio->req.state = STATE_IOREQ_NONE; break; } @@ -316,7 +316,7 @@ static int hvmemul_do_io( if ( dir == IOREQ_READ ) { rc = hvm_process_io_intercept(&ioreq_server_handler, &p); - vio->io_req.state = STATE_IOREQ_NONE; + vio->req.state = STATE_IOREQ_NONE; break; } } @@ -329,14 +329,14 @@ static int hvmemul_do_io( if ( !s ) { rc = hvm_process_io_intercept(&null_handler, &p); - vio->io_req.state = STATE_IOREQ_NONE; + vio->req.state = STATE_IOREQ_NONE; } else { rc = hvm_send_ioreq(s, &p, 0); if ( rc != X86EMUL_RETRY || currd->is_shutting_down ) - vio->io_req.state = STATE_IOREQ_NONE; - else if ( !ioreq_needs_completion(&vio->io_req) ) + vio->req.state = STATE_IOREQ_NONE; + else if ( !ioreq_needs_completion(&vio->req) ) rc = X86EMUL_OKAY; } break; @@ -1005,14 +1005,14 @@ static int hvmemul_phys_mmio_access( * cache indexed by linear MMIO address. */ static struct hvm_mmio_cache *hvmemul_find_mmio_cache( - struct hvm_vcpu_io *vio, unsigned long gla, uint8_t dir, bool create) + struct hvm_vcpu_io *hvio, unsigned long gla, uint8_t dir, bool create) { unsigned int i; struct hvm_mmio_cache *cache; - for ( i = 0; i < vio->mmio_cache_count; i ++ ) + for ( i = 0; i < hvio->mmio_cache_count; i ++ ) { - cache = &vio->mmio_cache[i]; + cache = &hvio->mmio_cache[i]; if ( gla == cache->gla && dir == cache->dir ) @@ -1022,13 +1022,13 @@ static struct hvm_mmio_cache *hvmemul_find_mmio_cache( if ( !create ) return NULL; - i = vio->mmio_cache_count; - if( i == ARRAY_SIZE(vio->mmio_cache) ) + i = hvio->mmio_cache_count; + if( i == ARRAY_SIZE(hvio->mmio_cache) ) return NULL; - ++vio->mmio_cache_count; + ++hvio->mmio_cache_count; - cache = &vio->mmio_cache[i]; + cache = &hvio->mmio_cache[i]; memset(cache, 0, sizeof (*cache)); cache->gla = gla; @@ -1037,26 +1037,26 @@ static struct hvm_mmio_cache *hvmemul_find_mmio_cache( return cache; } -static void latch_linear_to_phys(struct hvm_vcpu_io *vio, unsigned long gla, +static void latch_linear_to_phys(struct hvm_vcpu_io *hvio, unsigned long gla, unsigned long gpa, bool_t write) { - if ( vio->mmio_access.gla_valid ) + if ( hvio->mmio_access.gla_valid ) return; - vio->mmio_gla = gla & PAGE_MASK; - vio->mmio_gpfn = PFN_DOWN(gpa); - vio->mmio_access = (struct npfec){ .gla_valid = 1, - .read_access = 1, - .write_access = write }; + hvio->mmio_gla = gla & PAGE_MASK; + hvio->mmio_gpfn = PFN_DOWN(gpa); + hvio->mmio_access = (struct npfec){ .gla_valid = 1, + .read_access = 1, + .write_access = write }; } static int hvmemul_linear_mmio_access( unsigned long gla, unsigned int size, uint8_t dir, void *buffer, uint32_t pfec, struct hvm_emulate_ctxt *hvmemul_ctxt, bool_t known_gpfn) { - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; unsigned long offset = gla & ~PAGE_MASK; - struct hvm_mmio_cache *cache = hvmemul_find_mmio_cache(vio, gla, dir, true); + struct hvm_mmio_cache *cache = hvmemul_find_mmio_cache(hvio, gla, dir, true); unsigned int chunk, buffer_offset = 0; paddr_t gpa; unsigned long one_rep = 1; @@ -1068,7 +1068,7 @@ static int hvmemul_linear_mmio_access( chunk = min_t(unsigned int, size, PAGE_SIZE - offset); if ( known_gpfn ) - gpa = pfn_to_paddr(vio->mmio_gpfn) | offset; + gpa = pfn_to_paddr(hvio->mmio_gpfn) | offset; else { rc = hvmemul_linear_to_phys(gla, &gpa, chunk, &one_rep, pfec, @@ -1076,7 +1076,7 @@ static int hvmemul_linear_mmio_access( if ( rc != X86EMUL_OKAY ) return rc; - latch_linear_to_phys(vio, gla, gpa, dir == IOREQ_WRITE); + latch_linear_to_phys(hvio, gla, gpa, dir == IOREQ_WRITE); } for ( ;; ) @@ -1122,22 +1122,22 @@ static inline int hvmemul_linear_mmio_write( static bool known_gla(unsigned long addr, unsigned int bytes, uint32_t pfec) { - const struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; + const struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; if ( pfec & PFEC_write_access ) { - if ( !vio->mmio_access.write_access ) + if ( !hvio->mmio_access.write_access ) return false; } else if ( pfec & PFEC_insn_fetch ) { - if ( !vio->mmio_access.insn_fetch ) + if ( !hvio->mmio_access.insn_fetch ) return false; } - else if ( !vio->mmio_access.read_access ) + else if ( !hvio->mmio_access.read_access ) return false; - return (vio->mmio_gla == (addr & PAGE_MASK) && + return (hvio->mmio_gla == (addr & PAGE_MASK) && (addr & ~PAGE_MASK) + bytes <= PAGE_SIZE); } @@ -1145,7 +1145,7 @@ static int linear_read(unsigned long addr, unsigned int bytes, void *p_data, uint32_t pfec, struct hvm_emulate_ctxt *hvmemul_ctxt) { pagefault_info_t pfinfo; - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; unsigned int offset = addr & ~PAGE_MASK; int rc = HVMTRANS_bad_gfn_to_mfn; @@ -1167,7 +1167,7 @@ static int linear_read(unsigned long addr, unsigned int bytes, void *p_data, * we handle this access in the same way to guarantee completion and hence * clean up any interim state. */ - if ( !hvmemul_find_mmio_cache(vio, addr, IOREQ_READ, false) ) + if ( !hvmemul_find_mmio_cache(hvio, addr, IOREQ_READ, false) ) rc = hvm_copy_from_guest_linear(p_data, addr, bytes, pfec, &pfinfo); switch ( rc ) @@ -1200,7 +1200,7 @@ static int linear_write(unsigned long addr, unsigned int bytes, void *p_data, uint32_t pfec, struct hvm_emulate_ctxt *hvmemul_ctxt) { pagefault_info_t pfinfo; - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; unsigned int offset = addr & ~PAGE_MASK; int rc = HVMTRANS_bad_gfn_to_mfn; @@ -1222,7 +1222,7 @@ static int linear_write(unsigned long addr, unsigned int bytes, void *p_data, * we handle this access in the same way to guarantee completion and hence * clean up any interim state. */ - if ( !hvmemul_find_mmio_cache(vio, addr, IOREQ_WRITE, false) ) + if ( !hvmemul_find_mmio_cache(hvio, addr, IOREQ_WRITE, false) ) rc = hvm_copy_to_guest_linear(addr, p_data, bytes, pfec, &pfinfo); switch ( rc ) @@ -1599,7 +1599,7 @@ static int hvmemul_cmpxchg( struct vcpu *curr = current; unsigned long addr; uint32_t pfec = PFEC_page_present | PFEC_write_access; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; int rc; void *mapping = NULL; @@ -1625,8 +1625,8 @@ static int hvmemul_cmpxchg( /* Fix this in case the guest is really relying on r-m-w atomicity. */ return hvmemul_linear_mmio_write(addr, bytes, p_new, pfec, hvmemul_ctxt, - vio->mmio_access.write_access && - vio->mmio_gla == (addr & PAGE_MASK)); + hvio->mmio_access.write_access && + hvio->mmio_gla == (addr & PAGE_MASK)); } switch ( bytes ) @@ -1823,7 +1823,7 @@ static int hvmemul_rep_movs( struct hvm_emulate_ctxt *hvmemul_ctxt = container_of(ctxt, struct hvm_emulate_ctxt, ctxt); struct vcpu *curr = current; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; unsigned long saddr, daddr, bytes; paddr_t sgpa, dgpa; uint32_t pfec = PFEC_page_present; @@ -1846,18 +1846,18 @@ static int hvmemul_rep_movs( if ( hvmemul_ctxt->seg_reg[x86_seg_ss].dpl == 3 ) pfec |= PFEC_user_mode; - if ( vio->mmio_access.read_access && - (vio->mmio_gla == (saddr & PAGE_MASK)) && + if ( hvio->mmio_access.read_access && + (hvio->mmio_gla == (saddr & PAGE_MASK)) && /* * Upon initial invocation don't truncate large batches just because * of a hit for the translation: Doing the guest page table walk is * cheaper than multiple round trips through the device model. Yet * when processing a response we can always re-use the translation. */ - (vio->io_req.state == STATE_IORESP_READY || + (curr->io.req.state == STATE_IORESP_READY || ((!df || *reps == 1) && PAGE_SIZE - (saddr & ~PAGE_MASK) >= *reps * bytes_per_rep)) ) - sgpa = pfn_to_paddr(vio->mmio_gpfn) | (saddr & ~PAGE_MASK); + sgpa = pfn_to_paddr(hvio->mmio_gpfn) | (saddr & ~PAGE_MASK); else { rc = hvmemul_linear_to_phys(saddr, &sgpa, bytes_per_rep, reps, pfec, @@ -1867,13 +1867,13 @@ static int hvmemul_rep_movs( } bytes = PAGE_SIZE - (daddr & ~PAGE_MASK); - if ( vio->mmio_access.write_access && - (vio->mmio_gla == (daddr & PAGE_MASK)) && + if ( hvio->mmio_access.write_access && + (hvio->mmio_gla == (daddr & PAGE_MASK)) && /* See comment above. */ - (vio->io_req.state == STATE_IORESP_READY || + (curr->io.req.state == STATE_IORESP_READY || ((!df || *reps == 1) && PAGE_SIZE - (daddr & ~PAGE_MASK) >= *reps * bytes_per_rep)) ) - dgpa = pfn_to_paddr(vio->mmio_gpfn) | (daddr & ~PAGE_MASK); + dgpa = pfn_to_paddr(hvio->mmio_gpfn) | (daddr & ~PAGE_MASK); else { rc = hvmemul_linear_to_phys(daddr, &dgpa, bytes_per_rep, reps, @@ -1892,14 +1892,14 @@ static int hvmemul_rep_movs( if ( sp2mt == p2m_mmio_dm ) { - latch_linear_to_phys(vio, saddr, sgpa, 0); + latch_linear_to_phys(hvio, saddr, sgpa, 0); return hvmemul_do_mmio_addr( sgpa, reps, bytes_per_rep, IOREQ_READ, df, dgpa); } if ( dp2mt == p2m_mmio_dm ) { - latch_linear_to_phys(vio, daddr, dgpa, 1); + latch_linear_to_phys(hvio, daddr, dgpa, 1); return hvmemul_do_mmio_addr( dgpa, reps, bytes_per_rep, IOREQ_WRITE, df, sgpa); } @@ -1992,7 +1992,7 @@ static int hvmemul_rep_stos( struct hvm_emulate_ctxt *hvmemul_ctxt = container_of(ctxt, struct hvm_emulate_ctxt, ctxt); struct vcpu *curr = current; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; unsigned long addr, bytes; paddr_t gpa; p2m_type_t p2mt; @@ -2004,13 +2004,13 @@ static int hvmemul_rep_stos( return rc; bytes = PAGE_SIZE - (addr & ~PAGE_MASK); - if ( vio->mmio_access.write_access && - (vio->mmio_gla == (addr & PAGE_MASK)) && + if ( hvio->mmio_access.write_access && + (hvio->mmio_gla == (addr & PAGE_MASK)) && /* See respective comment in MOVS processing. */ - (vio->io_req.state == STATE_IORESP_READY || + (curr->io.req.state == STATE_IORESP_READY || ((!df || *reps == 1) && PAGE_SIZE - (addr & ~PAGE_MASK) >= *reps * bytes_per_rep)) ) - gpa = pfn_to_paddr(vio->mmio_gpfn) | (addr & ~PAGE_MASK); + gpa = pfn_to_paddr(hvio->mmio_gpfn) | (addr & ~PAGE_MASK); else { uint32_t pfec = PFEC_page_present | PFEC_write_access; @@ -2103,7 +2103,7 @@ static int hvmemul_rep_stos( return X86EMUL_UNHANDLEABLE; case p2m_mmio_dm: - latch_linear_to_phys(vio, addr, gpa, 1); + latch_linear_to_phys(hvio, addr, gpa, 1); return hvmemul_do_mmio_buffer(gpa, reps, bytes_per_rep, IOREQ_WRITE, df, p_data); } @@ -2613,18 +2613,18 @@ static const struct x86_emulate_ops hvm_emulate_ops_no_write = { }; /* - * Note that passing HVMIO_no_completion into this function serves as kind + * Note that passing VIO_no_completion into this function serves as kind * of (but not fully) an "auto select completion" indicator. When there's * no completion needed, the passed in value will be ignored in any case. */ static int _hvm_emulate_one(struct hvm_emulate_ctxt *hvmemul_ctxt, const struct x86_emulate_ops *ops, - enum hvm_io_completion completion) + enum vio_completion completion) { const struct cpu_user_regs *regs = hvmemul_ctxt->ctxt.regs; struct vcpu *curr = current; uint32_t new_intr_shadow; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; int rc; /* @@ -2632,45 +2632,45 @@ static int _hvm_emulate_one(struct hvm_emulate_ctxt *hvmemul_ctxt, * untouched if it's already enabled, for re-execution to consume * entries populated by an earlier pass. */ - if ( vio->cache->num_ents > vio->cache->max_ents ) + if ( hvio->cache->num_ents > hvio->cache->max_ents ) { - ASSERT(vio->io_req.state == STATE_IOREQ_NONE); - vio->cache->num_ents = 0; + ASSERT(curr->io.req.state == STATE_IOREQ_NONE); + hvio->cache->num_ents = 0; } else - ASSERT(vio->io_req.state == STATE_IORESP_READY); + ASSERT(curr->io.req.state == STATE_IORESP_READY); - hvm_emulate_init_per_insn(hvmemul_ctxt, vio->mmio_insn, - vio->mmio_insn_bytes); + hvm_emulate_init_per_insn(hvmemul_ctxt, hvio->mmio_insn, + hvio->mmio_insn_bytes); - vio->mmio_retry = 0; + hvio->mmio_retry = 0; rc = x86_emulate(&hvmemul_ctxt->ctxt, ops); - if ( rc == X86EMUL_OKAY && vio->mmio_retry ) + if ( rc == X86EMUL_OKAY && hvio->mmio_retry ) rc = X86EMUL_RETRY; - if ( !ioreq_needs_completion(&vio->io_req) ) - completion = HVMIO_no_completion; - else if ( completion == HVMIO_no_completion ) - completion = (vio->io_req.type != IOREQ_TYPE_PIO || - hvmemul_ctxt->is_mem_access) ? HVMIO_mmio_completion - : HVMIO_pio_completion; + if ( !ioreq_needs_completion(&curr->io.req) ) + completion = VIO_no_completion; + else if ( completion == VIO_no_completion ) + completion = (curr->io.req.type != IOREQ_TYPE_PIO || + hvmemul_ctxt->is_mem_access) ? VIO_mmio_completion + : VIO_pio_completion; - switch ( vio->io_completion = completion ) + switch ( curr->io.completion = completion ) { - case HVMIO_no_completion: - case HVMIO_pio_completion: - vio->mmio_cache_count = 0; - vio->mmio_insn_bytes = 0; - vio->mmio_access = (struct npfec){}; + case VIO_no_completion: + case VIO_pio_completion: + hvio->mmio_cache_count = 0; + hvio->mmio_insn_bytes = 0; + hvio->mmio_access = (struct npfec){}; hvmemul_cache_disable(curr); break; - case HVMIO_mmio_completion: - case HVMIO_realmode_completion: - BUILD_BUG_ON(sizeof(vio->mmio_insn) < sizeof(hvmemul_ctxt->insn_buf)); - vio->mmio_insn_bytes = hvmemul_ctxt->insn_buf_bytes; - memcpy(vio->mmio_insn, hvmemul_ctxt->insn_buf, vio->mmio_insn_bytes); + case VIO_mmio_completion: + case VIO_realmode_completion: + BUILD_BUG_ON(sizeof(hvio->mmio_insn) < sizeof(hvmemul_ctxt->insn_buf)); + hvio->mmio_insn_bytes = hvmemul_ctxt->insn_buf_bytes; + memcpy(hvio->mmio_insn, hvmemul_ctxt->insn_buf, hvio->mmio_insn_bytes); break; default: @@ -2716,7 +2716,7 @@ static int _hvm_emulate_one(struct hvm_emulate_ctxt *hvmemul_ctxt, int hvm_emulate_one( struct hvm_emulate_ctxt *hvmemul_ctxt, - enum hvm_io_completion completion) + enum vio_completion completion) { return _hvm_emulate_one(hvmemul_ctxt, &hvm_emulate_ops, completion); } @@ -2754,7 +2754,7 @@ int hvm_emulate_one_mmio(unsigned long mfn, unsigned long gla) guest_cpu_user_regs()); ctxt.ctxt.data = &mmio_ro_ctxt; - switch ( rc = _hvm_emulate_one(&ctxt, ops, HVMIO_no_completion) ) + switch ( rc = _hvm_emulate_one(&ctxt, ops, VIO_no_completion) ) { case X86EMUL_UNHANDLEABLE: case X86EMUL_UNIMPLEMENTED: @@ -2782,28 +2782,28 @@ void hvm_emulate_one_vm_event(enum emul_kind kind, unsigned int trapnr, { case EMUL_KIND_NOWRITE: rc = _hvm_emulate_one(&ctx, &hvm_emulate_ops_no_write, - HVMIO_no_completion); + VIO_no_completion); break; case EMUL_KIND_SET_CONTEXT_INSN: { struct vcpu *curr = current; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; - BUILD_BUG_ON(sizeof(vio->mmio_insn) != + BUILD_BUG_ON(sizeof(hvio->mmio_insn) != sizeof(curr->arch.vm_event->emul.insn.data)); - ASSERT(!vio->mmio_insn_bytes); + ASSERT(!hvio->mmio_insn_bytes); /* * Stash insn buffer into mmio buffer here instead of ctx * to avoid having to add more logic to hvm_emulate_one. */ - vio->mmio_insn_bytes = sizeof(vio->mmio_insn); - memcpy(vio->mmio_insn, curr->arch.vm_event->emul.insn.data, - vio->mmio_insn_bytes); + hvio->mmio_insn_bytes = sizeof(hvio->mmio_insn); + memcpy(hvio->mmio_insn, curr->arch.vm_event->emul.insn.data, + hvio->mmio_insn_bytes); } /* Fall-through */ default: ctx.set_context = (kind == EMUL_KIND_SET_CONTEXT_DATA); - rc = hvm_emulate_one(&ctx, HVMIO_no_completion); + rc = hvm_emulate_one(&ctx, VIO_no_completion); } switch ( rc ) diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c index bc96947..4ed929c 100644 --- a/xen/arch/x86/hvm/hvm.c +++ b/xen/arch/x86/hvm/hvm.c @@ -3800,7 +3800,7 @@ void hvm_ud_intercept(struct cpu_user_regs *regs) return; } - switch ( hvm_emulate_one(&ctxt, HVMIO_no_completion) ) + switch ( hvm_emulate_one(&ctxt, VIO_no_completion) ) { case X86EMUL_UNHANDLEABLE: case X86EMUL_UNIMPLEMENTED: diff --git a/xen/arch/x86/hvm/io.c b/xen/arch/x86/hvm/io.c index ef8286b..dd733e1 100644 --- a/xen/arch/x86/hvm/io.c +++ b/xen/arch/x86/hvm/io.c @@ -85,7 +85,7 @@ bool hvm_emulate_one_insn(hvm_emulate_validate_t *validate, const char *descr) hvm_emulate_init_once(&ctxt, validate, guest_cpu_user_regs()); - switch ( rc = hvm_emulate_one(&ctxt, HVMIO_no_completion) ) + switch ( rc = hvm_emulate_one(&ctxt, VIO_no_completion) ) { case X86EMUL_UNHANDLEABLE: hvm_dump_emulation_state(XENLOG_G_WARNING, descr, &ctxt, rc); @@ -109,20 +109,20 @@ bool hvm_emulate_one_insn(hvm_emulate_validate_t *validate, const char *descr) bool handle_mmio_with_translation(unsigned long gla, unsigned long gpfn, struct npfec access) { - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; - vio->mmio_access = access.gla_valid && - access.kind == npfec_kind_with_gla - ? access : (struct npfec){}; - vio->mmio_gla = gla & PAGE_MASK; - vio->mmio_gpfn = gpfn; + hvio->mmio_access = access.gla_valid && + access.kind == npfec_kind_with_gla + ? access : (struct npfec){}; + hvio->mmio_gla = gla & PAGE_MASK; + hvio->mmio_gpfn = gpfn; return handle_mmio(); } bool handle_pio(uint16_t port, unsigned int size, int dir) { struct vcpu *curr = current; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct vcpu_io *vio = &curr->io; unsigned int data; int rc; @@ -135,8 +135,8 @@ bool handle_pio(uint16_t port, unsigned int size, int dir) rc = hvmemul_do_pio_buffer(port, size, dir, &data); - if ( ioreq_needs_completion(&vio->io_req) ) - vio->io_completion = HVMIO_pio_completion; + if ( ioreq_needs_completion(&vio->req) ) + vio->completion = VIO_pio_completion; switch ( rc ) { @@ -175,7 +175,7 @@ static bool_t g2m_portio_accept(const struct hvm_io_handler *handler, { struct vcpu *curr = current; const struct hvm_domain *hvm = &curr->domain->arch.hvm; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; struct g2m_ioport *g2m_ioport; unsigned int start, end; @@ -185,7 +185,7 @@ static bool_t g2m_portio_accept(const struct hvm_io_handler *handler, end = start + g2m_ioport->np; if ( (p->addr >= start) && (p->addr + p->size <= end) ) { - vio->g2m_ioport = g2m_ioport; + hvio->g2m_ioport = g2m_ioport; return 1; } } @@ -196,8 +196,8 @@ static bool_t g2m_portio_accept(const struct hvm_io_handler *handler, static int g2m_portio_read(const struct hvm_io_handler *handler, uint64_t addr, uint32_t size, uint64_t *data) { - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; - const struct g2m_ioport *g2m_ioport = vio->g2m_ioport; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; + const struct g2m_ioport *g2m_ioport = hvio->g2m_ioport; unsigned int mport = (addr - g2m_ioport->gport) + g2m_ioport->mport; switch ( size ) @@ -221,8 +221,8 @@ static int g2m_portio_read(const struct hvm_io_handler *handler, static int g2m_portio_write(const struct hvm_io_handler *handler, uint64_t addr, uint32_t size, uint64_t data) { - struct hvm_vcpu_io *vio = ¤t->arch.hvm.hvm_io; - const struct g2m_ioport *g2m_ioport = vio->g2m_ioport; + struct hvm_vcpu_io *hvio = ¤t->arch.hvm.hvm_io; + const struct g2m_ioport *g2m_ioport = hvio->g2m_ioport; unsigned int mport = (addr - g2m_ioport->gport) + g2m_ioport->mport; switch ( size ) diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 0cadf34..62a4b33 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -40,11 +40,11 @@ bool arch_ioreq_complete_mmio(void) return handle_mmio(); } -bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion) +bool arch_vcpu_ioreq_completion(enum vio_completion completion) { - switch ( io_completion ) + switch ( completion ) { - case HVMIO_realmode_completion: + case VIO_realmode_completion: { struct hvm_emulate_ctxt ctxt; diff --git a/xen/arch/x86/hvm/svm/nestedsvm.c b/xen/arch/x86/hvm/svm/nestedsvm.c index fcfccf7..6d90630 100644 --- a/xen/arch/x86/hvm/svm/nestedsvm.c +++ b/xen/arch/x86/hvm/svm/nestedsvm.c @@ -1266,7 +1266,7 @@ enum hvm_intblk nsvm_intr_blocked(struct vcpu *v) * Delay the injection because this would result in delivering * an interrupt *within* the execution of an instruction. */ - if ( v->arch.hvm.hvm_io.io_req.state != STATE_IOREQ_NONE ) + if ( v->io.req.state != STATE_IOREQ_NONE ) return hvm_intblk_shadow; if ( !nv->nv_vmexit_pending && n2vmcb->exit_int_info.v ) diff --git a/xen/arch/x86/hvm/vmx/realmode.c b/xen/arch/x86/hvm/vmx/realmode.c index 768f01e..cc23afa 100644 --- a/xen/arch/x86/hvm/vmx/realmode.c +++ b/xen/arch/x86/hvm/vmx/realmode.c @@ -101,7 +101,7 @@ void vmx_realmode_emulate_one(struct hvm_emulate_ctxt *hvmemul_ctxt) perfc_incr(realmode_emulations); - rc = hvm_emulate_one(hvmemul_ctxt, HVMIO_realmode_completion); + rc = hvm_emulate_one(hvmemul_ctxt, VIO_realmode_completion); if ( rc == X86EMUL_UNHANDLEABLE ) { @@ -153,7 +153,7 @@ void vmx_realmode(struct cpu_user_regs *regs) struct vcpu *curr = current; struct hvm_emulate_ctxt hvmemul_ctxt; struct segment_register *sreg; - struct hvm_vcpu_io *vio = &curr->arch.hvm.hvm_io; + struct hvm_vcpu_io *hvio = &curr->arch.hvm.hvm_io; unsigned long intr_info; unsigned int emulations = 0; @@ -188,7 +188,7 @@ void vmx_realmode(struct cpu_user_regs *regs) vmx_realmode_emulate_one(&hvmemul_ctxt); - if ( vio->io_req.state != STATE_IOREQ_NONE || vio->mmio_retry ) + if ( curr->io.req.state != STATE_IOREQ_NONE || hvio->mmio_retry ) break; /* Stop emulating unless our segment state is not safe */ @@ -202,7 +202,7 @@ void vmx_realmode(struct cpu_user_regs *regs) } /* Need to emulate next time if we've started an IO operation */ - if ( vio->io_req.state != STATE_IOREQ_NONE ) + if ( curr->io.req.state != STATE_IOREQ_NONE ) curr->arch.hvm.vmx.vmx_emulate = 1; if ( !curr->arch.hvm.vmx.vmx_emulate && !curr->arch.hvm.vmx.vmx_realmode ) diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 84bce36..ce3ef59 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -159,7 +159,7 @@ static bool hvm_wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) break; } - p = &sv->vcpu->arch.hvm.hvm_io.io_req; + p = &sv->vcpu->io.req; if ( ioreq_needs_completion(p) ) p->data = data; @@ -171,10 +171,10 @@ static bool hvm_wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) bool handle_hvm_io_completion(struct vcpu *v) { struct domain *d = v->domain; - struct hvm_vcpu_io *vio = &v->arch.hvm.hvm_io; + struct vcpu_io *vio = &v->io; struct ioreq_server *s; struct ioreq_vcpu *sv; - enum hvm_io_completion io_completion; + enum vio_completion completion; if ( has_vpci(d) && vpci_process_pending(v) ) { @@ -186,29 +186,29 @@ bool handle_hvm_io_completion(struct vcpu *v) if ( sv && !hvm_wait_for_io(sv, get_ioreq(s, v)) ) return false; - vio->io_req.state = ioreq_needs_completion(&vio->io_req) ? + vio->req.state = ioreq_needs_completion(&vio->req) ? STATE_IORESP_READY : STATE_IOREQ_NONE; msix_write_completion(v); vcpu_end_shutdown_deferral(v); - io_completion = vio->io_completion; - vio->io_completion = HVMIO_no_completion; + completion = vio->completion; + vio->completion = VIO_no_completion; - switch ( io_completion ) + switch ( completion ) { - case HVMIO_no_completion: + case VIO_no_completion: break; - case HVMIO_mmio_completion: + case VIO_mmio_completion: return arch_ioreq_complete_mmio(); - case HVMIO_pio_completion: - return handle_pio(vio->io_req.addr, vio->io_req.size, - vio->io_req.dir); + case VIO_pio_completion: + return handle_pio(vio->req.addr, vio->req.size, + vio->req.dir); default: - return arch_vcpu_ioreq_completion(io_completion); + return arch_vcpu_ioreq_completion(completion); } return true; diff --git a/xen/include/asm-x86/hvm/emulate.h b/xen/include/asm-x86/hvm/emulate.h index 1620cc7..610078b 100644 --- a/xen/include/asm-x86/hvm/emulate.h +++ b/xen/include/asm-x86/hvm/emulate.h @@ -65,7 +65,7 @@ bool __nonnull(1, 2) hvm_emulate_one_insn( const char *descr); int hvm_emulate_one( struct hvm_emulate_ctxt *hvmemul_ctxt, - enum hvm_io_completion completion); + enum vio_completion completion); void hvm_emulate_one_vm_event(enum emul_kind kind, unsigned int trapnr, unsigned int errcode); diff --git a/xen/include/asm-x86/hvm/vcpu.h b/xen/include/asm-x86/hvm/vcpu.h index 6c1feda..8adf455 100644 --- a/xen/include/asm-x86/hvm/vcpu.h +++ b/xen/include/asm-x86/hvm/vcpu.h @@ -28,13 +28,6 @@ #include #include -enum hvm_io_completion { - HVMIO_no_completion, - HVMIO_mmio_completion, - HVMIO_pio_completion, - HVMIO_realmode_completion -}; - struct hvm_vcpu_asid { uint64_t generation; uint32_t asid; @@ -52,10 +45,6 @@ struct hvm_mmio_cache { }; struct hvm_vcpu_io { - /* I/O request in flight to device model. */ - enum hvm_io_completion io_completion; - ioreq_t io_req; - /* * HVM emulation: * Linear address @mmio_gla maps to MMIO physical frame @mmio_gpfn. diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 60e864d..eace1d3 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -107,7 +107,7 @@ void hvm_ioreq_init(struct domain *d); int ioreq_server_dm_op(struct xen_dm_op *op, struct domain *d, bool *const_op); bool arch_ioreq_complete_mmio(void); -bool arch_vcpu_ioreq_completion(enum hvm_io_completion io_completion); +bool arch_vcpu_ioreq_completion(enum vio_completion completion); int arch_ioreq_server_map_pages(struct ioreq_server *s); void arch_ioreq_server_unmap_pages(struct ioreq_server *s); void arch_ioreq_server_enable(struct ioreq_server *s); diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index f437ee3..59e5b6a 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -147,6 +147,21 @@ void evtchn_destroy_final(struct domain *d); /* from complete_domain_destroy */ struct waitqueue_vcpu; +enum vio_completion { + VIO_no_completion, + VIO_mmio_completion, + VIO_pio_completion, +#ifdef CONFIG_X86 + VIO_realmode_completion, +#endif +}; + +struct vcpu_io { + /* I/O request in flight to device model. */ + enum vio_completion completion; + ioreq_t req; +}; + struct vcpu { int vcpu_id; @@ -258,6 +273,10 @@ struct vcpu struct vpci_vcpu vpci; struct arch_vcpu arch; + +#ifdef CONFIG_IOREQ_SERVER + struct vcpu_io io; +#endif }; struct sched_unit { From patchwork Mon Jan 25 19:08:18 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044091 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id A84C4C433E0 for ; Mon, 25 Jan 2021 19:20:12 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 4A1EE20E65 for ; Mon, 25 Jan 2021 19:20:12 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 4A1EE20E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74354.133690 (Exim 4.92) (envelope-from ) id 1l47P8-0002Ft-Sp; Mon, 25 Jan 2021 19:20:02 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74354.133690; Mon, 25 Jan 2021 19:20:02 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P8-0002Ex-G9; Mon, 25 Jan 2021 19:20:02 +0000 Received: by outflank-mailman (input) for mailman id 74354; Mon, 25 Jan 2021 19:20:00 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47FG-0008N7-Ik for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:09:50 +0000 Received: from mail-wm1-x32a.google.com (unknown [2a00:1450:4864:20::32a]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 54e1574d-0cff-4cd0-a3d2-4f4dfbd2f06b; Mon, 25 Jan 2021 19:09:05 +0000 (UTC) Received: by mail-wm1-x32a.google.com with SMTP id 190so404718wmz.0 for ; Mon, 25 Jan 2021 11:09:05 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.03 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:04 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 54e1574d-0cff-4cd0-a3d2-4f4dfbd2f06b DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=XJ5PwD+gBkmZaN/B9jSk8lYAoHZ9OkzGDLF+hASZ6yI=; b=KmhCDvfatRhc71tGFTQX4nkHChngGSENcsKJt2izIcVCIV6Am/SDrJwD4AR/uKJ1xx DbWCugw4ho8Yj1zdkeJOLlImxKGUgX+tDnEPedcWbluhUgKmJLa7PXsdAodx+6bjRNvt mqqJQUXmXpNpdhcC/y0sSViD2SUxdYa+iqudDRTMWfAgLvN8MCDC8gIKOodptQKAP52B GzGPtqbnZIXFWzgW1J1NWs//cTMO2Y3IF+kDAmcmQ9Mn+8VbthrE/BhxVydjkwBojZ4i 1FdspAH4dgq4aBMtWwqpYoQtGuPXJSv1/Z9PYUd55Qm2Hv+3m7CO9A3sQnMNWZnt0BqY VXyA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=XJ5PwD+gBkmZaN/B9jSk8lYAoHZ9OkzGDLF+hASZ6yI=; b=F9/en+xAkNWR4omFYFytu4ZQilEjTw50Jjj5IVc1ePlP2agvhhmz6juDY6qsPCjmbh Cu3K5nRj0SHOmSt2Q07j+GTbg/iv2X3l1++nWSu8Y8w5Wq+5VHmrOjkWCA4S8hbqNgE9 XS19ObgVNnmpcgJAI3NHKnOimfW3F7nsL7HzYDr/1dPpL3IlwjAV67b9uFgvOMHNuuNA ZoiKxNiMNW9/JtdNsyVCJ7ys6RdGr3jdgbNre7hcw38TcFht1xUGW6uceitxebwKw1Ce mE63rCZ/7DMbcluTXpeZ4JzWDSStfPJeVlEnRixTG4XtQZGmuV0rLF3xxOX88KVa3FP+ D4Gg== X-Gm-Message-State: AOAM533cMg+cJkgsi7XQwY0oTgtL8GQNzoxOWeTLPVeRf5/MWsnnkWP3 cjOx6MJnggZ0m4zU0ka+ZFTqTULHm1aUzg== X-Google-Smtp-Source: ABdhPJzmnPli3lOGk5X8VdadTB+eEwmbGIbFMGNRuUcrEoqH3VBWstVH3lJmMEyD9osn3jI+Dr7MQw== X-Received: by 2002:a05:600c:21d8:: with SMTP id x24mr1443871wmj.27.1611601744646; Mon, 25 Jan 2021 11:09:04 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Julien Grall , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Ian Jackson , Julien Grall , Stefano Stabellini , Oleksandr Tyshchenko Subject: [PATCH V5 11/22] xen/mm: Make x86's XENMEM_resource_ioreq_server handling common Date: Mon, 25 Jan 2021 21:08:18 +0200 Message-Id: <1611601709-28361-12-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Julien Grall As x86 implementation of XENMEM_resource_ioreq_server can be re-used on Arm later on, this patch makes it common and removes arch_acquire_resource (and the corresponding option) as unneeded. Also re-order #include-s alphabetically. This support is going to be used on Arm to be able run device emulator outside of Xen hypervisor. Signed-off-by: Julien Grall Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Jan Beulich Reviewed-by: Paul Durrant [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - no changes Changes V1 -> V2: - update the author of a patch Changes V2 -> V3: - don't wrap #include - limit the number of #ifdef-s - re-order #include-s alphabetically Changes V3 -> V4: - rebase - Add Jan's R-b Changes V4 -> V5: - add Paul's R-b - update patch description - remove ARCH_ACQUIRE_RESOURCE option, etc --- --- xen/arch/x86/Kconfig | 1 - xen/arch/x86/mm.c | 44 --------------------------------- xen/common/Kconfig | 3 --- xen/common/memory.c | 63 +++++++++++++++++++++++++++++++++++++++--------- xen/include/asm-x86/mm.h | 4 --- xen/include/xen/mm.h | 9 ------- 6 files changed, 51 insertions(+), 73 deletions(-) diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig index ea9a9ea..abe0fce 100644 --- a/xen/arch/x86/Kconfig +++ b/xen/arch/x86/Kconfig @@ -6,7 +6,6 @@ config X86 select ACPI select ACPI_LEGACY_TABLES_LOOKUP select ARCH_SUPPORTS_INT128 - select ARCH_ACQUIRE_RESOURCE select COMPAT select CORE_PARKING select HAS_ALTERNATIVE diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c index 59eb5c8..4366ea3 100644 --- a/xen/arch/x86/mm.c +++ b/xen/arch/x86/mm.c @@ -4587,50 +4587,6 @@ static int handle_iomem_range(unsigned long s, unsigned long e, void *p) return err || s > e ? err : _handle_iomem_range(s, e, p); } -int arch_acquire_resource(struct domain *d, unsigned int type, - unsigned int id, unsigned long frame, - unsigned int nr_frames, xen_pfn_t mfn_list[]) -{ - int rc; - - switch ( type ) - { -#ifdef CONFIG_HVM - case XENMEM_resource_ioreq_server: - { - ioservid_t ioservid = id; - unsigned int i; - - rc = -EINVAL; - if ( !is_hvm_domain(d) ) - break; - - if ( id != (unsigned int)ioservid ) - break; - - rc = 0; - for ( i = 0; i < nr_frames; i++ ) - { - mfn_t mfn; - - rc = hvm_get_ioreq_server_frame(d, id, frame + i, &mfn); - if ( rc ) - break; - - mfn_list[i] = mfn_x(mfn); - } - break; - } -#endif - - default: - rc = -EOPNOTSUPP; - break; - } - - return rc; -} - long arch_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) { int rc; diff --git a/xen/common/Kconfig b/xen/common/Kconfig index cf32a07..fa049a6 100644 --- a/xen/common/Kconfig +++ b/xen/common/Kconfig @@ -22,9 +22,6 @@ config GRANT_TABLE If unsure, say Y. -config ARCH_ACQUIRE_RESOURCE - bool - config HAS_ALTERNATIVE bool diff --git a/xen/common/memory.c b/xen/common/memory.c index ccb4d49..2f274a6 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -8,22 +8,23 @@ */ #include -#include +#include +#include +#include +#include +#include +#include +#include #include +#include #include +#include +#include #include #include #include -#include -#include -#include -#include -#include -#include -#include -#include #include -#include +#include #include #include #include @@ -1091,6 +1092,40 @@ static int acquire_grant_table(struct domain *d, unsigned int id, return 0; } +static int acquire_ioreq_server(struct domain *d, + unsigned int id, + unsigned long frame, + unsigned int nr_frames, + xen_pfn_t mfn_list[]) +{ +#ifdef CONFIG_IOREQ_SERVER + ioservid_t ioservid = id; + unsigned int i; + int rc; + + if ( !is_hvm_domain(d) ) + return -EINVAL; + + if ( id != (unsigned int)ioservid ) + return -EINVAL; + + for ( i = 0; i < nr_frames; i++ ) + { + mfn_t mfn; + + rc = hvm_get_ioreq_server_frame(d, id, frame + i, &mfn); + if ( rc ) + return rc; + + mfn_list[i] = mfn_x(mfn); + } + + return 0; +#else + return -EOPNOTSUPP; +#endif +} + static int acquire_resource( XEN_GUEST_HANDLE_PARAM(xen_mem_acquire_resource_t) arg) { @@ -1149,9 +1184,13 @@ static int acquire_resource( mfn_list); break; + case XENMEM_resource_ioreq_server: + rc = acquire_ioreq_server(d, xmar.id, xmar.frame, xmar.nr_frames, + mfn_list); + break; + default: - rc = arch_acquire_resource(d, xmar.type, xmar.id, xmar.frame, - xmar.nr_frames, mfn_list); + rc = -EOPNOTSUPP; break; } diff --git a/xen/include/asm-x86/mm.h b/xen/include/asm-x86/mm.h index 1fdb4eb..041c158 100644 --- a/xen/include/asm-x86/mm.h +++ b/xen/include/asm-x86/mm.h @@ -640,8 +640,4 @@ static inline bool arch_mfn_in_directmap(unsigned long mfn) return mfn <= (virt_to_mfn(eva - 1) + 1); } -int arch_acquire_resource(struct domain *d, unsigned int type, - unsigned int id, unsigned long frame, - unsigned int nr_frames, xen_pfn_t mfn_list[]); - #endif /* __ASM_X86_MM_H__ */ diff --git a/xen/include/xen/mm.h b/xen/include/xen/mm.h index 636a125..667f9da 100644 --- a/xen/include/xen/mm.h +++ b/xen/include/xen/mm.h @@ -612,13 +612,4 @@ static inline void put_page_alloc_ref(struct page_info *page) } } -#ifndef CONFIG_ARCH_ACQUIRE_RESOURCE -static inline int arch_acquire_resource( - struct domain *d, unsigned int type, unsigned int id, unsigned long frame, - unsigned int nr_frames, xen_pfn_t mfn_list[]) -{ - return -EOPNOTSUPP; -} -#endif /* !CONFIG_ARCH_ACQUIRE_RESOURCE */ - #endif /* __XEN_MM_H__ */ From patchwork Mon Jan 25 19:08:19 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044081 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 07CE9C433E6 for ; Mon, 25 Jan 2021 19:20:07 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 9EF8720E65 for ; Mon, 25 Jan 2021 19:20:06 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 9EF8720E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74347.133632 (Exim 4.92) (envelope-from ) id 1l47P4-000251-Da; Mon, 25 Jan 2021 19:19:58 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74347.133632; Mon, 25 Jan 2021 19:19:58 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P4-00024n-AQ; Mon, 25 Jan 2021 19:19:58 +0000 Received: by outflank-mailman (input) for mailman id 74347; Mon, 25 Jan 2021 19:19:57 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47FQ-0008N7-HW for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:00 +0000 Received: from mail-wr1-x433.google.com (unknown [2a00:1450:4864:20::433]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id d990fcf5-f024-49c8-9935-8fe8c33f3b79; Mon, 25 Jan 2021 19:09:07 +0000 (UTC) Received: by mail-wr1-x433.google.com with SMTP id a1so14113609wrq.6 for ; Mon, 25 Jan 2021 11:09:07 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.04 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:05 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d990fcf5-f024-49c8-9935-8fe8c33f3b79 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=tBnvjjf4uUabJzL4II6Dn/93FRVBNLzjSNQLKAwEJvE=; b=E93exBdqqT8b1JrQ6VQUkZqVT2jisv52oG+CZbHFbBt+SeeFk3Zn7c1pJrQ+djk3ni CcrQqxEmm/NJU/XhhVfAa+7F/B7COpRoQtwAdKwbFSHBfs5uu8JdrYQx9Gih0fUY/KIQ XNm3Re0iB2SX4HImsviJWyZiHS0NMCrfDAMf/NuHfNTsdGMCOsA8i3h0KiStaGPNcn// Wqwy529Zb8pS5DrQp6NKpm0klgqkLk8+oqnZEMgYBcPQE+MkSVlIPVIoPoESuSB1oIuX Jd+9cFHBQcr6/lFlrwcxMC3SmEdv40JT385Kl0j57T3yhNSOBgqApY/6uxuI4pOdgmPc mcXA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=tBnvjjf4uUabJzL4II6Dn/93FRVBNLzjSNQLKAwEJvE=; b=lfqOz6OQPwKvjq7uPCuBRc64CR0Ck7l9jWrzjX0ZEZ8oxgnBnajebygZW5dIVDAU8S 4sGfn06ktZPvGw9dsV5XnKn9VByMqxVDFam5JqR/9dG1bmw+oSQI0S0KZ+QSNivXOkKt d4c5GP5DczTZwKqnZ3WLaDRoeZrtl5hCwDnmucGE3U5LKxU5yOTQAPO/+zMHiSyVJ2ye MTrunEirtQVuvvMVhe5groN51eQwR+nIs2j/ihCGKu+nO1SgdELUwAA22RMQk6ASSrGT gTfwwhaVVZ0Gp5uXTxmkd7oaFgqrPXkA7PAyhvMVKwTpmc8TJvtyVYY8TyMwFpQ+jwhm 4V1w== X-Gm-Message-State: AOAM532z4kD8NVtIAojjuSi/xxh7yviqASmsPFrjjDkoyEE2bp1uAZ7J BhtgDy/Alq7pox9gBRoiMPljYspSfGca0A== X-Google-Smtp-Source: ABdhPJx3YWttabJzb4Sa+Dem7efV32asMzEomZoW8vPVLsaQZDfXkwDcAnnKJddpd1489H1O0CmCTA== X-Received: by 2002:adf:8145:: with SMTP id 63mr2523046wrm.8.1611601745930; Mon, 25 Jan 2021 11:09:05 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Ian Jackson , Julien Grall , Stefano Stabellini , Jun Nakajima , Kevin Tian , Julien Grall Subject: [PATCH V5 12/22] xen/ioreq: Remove "hvm" prefixes from involved function names Date: Mon, 25 Jan 2021 21:08:19 +0200 Message-Id: <1611601709-28361-13-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko This patch removes "hvm" prefixes and infixes from IOREQ related function names in the common code and performs a renaming where appropriate according to the more consistent new naming scheme: - IOREQ server functions should start with "ioreq_server_" - IOREQ functions should start with "ioreq_" A few function names are clarified to better fit into their purposes: handle_hvm_io_completion -> vcpu_ioreq_handle_completion hvm_io_pending -> vcpu_ioreq_pending hvm_ioreq_init -> ioreq_domain_init hvm_alloc_ioreq_mfn -> ioreq_server_alloc_mfn hvm_free_ioreq_mfn -> ioreq_server_free_mfn Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Jan Beulich Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V1 -> V2: - new patch Changes V2 -> V3: - update patch according the "legacy interface" is x86 specific - update patch description - rename everything touched according to new naming scheme Changes V3 -> V4: - rebase - rename ioreq_update_evtchn() to ioreq_server_update_evtchn() - add Jan's R-b Changes V4 -> V5: - rebase - add Paul's R-b --- --- xen/arch/x86/hvm/emulate.c | 6 +- xen/arch/x86/hvm/hvm.c | 10 ++-- xen/arch/x86/hvm/io.c | 6 +- xen/arch/x86/hvm/ioreq.c | 2 +- xen/arch/x86/hvm/stdvga.c | 4 +- xen/arch/x86/hvm/vmx/vvmx.c | 2 +- xen/common/ioreq.c | 138 ++++++++++++++++++++++---------------------- xen/common/memory.c | 2 +- xen/include/xen/ioreq.h | 26 ++++----- 9 files changed, 98 insertions(+), 98 deletions(-) diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c index 21051ce..425c8dd 100644 --- a/xen/arch/x86/hvm/emulate.c +++ b/xen/arch/x86/hvm/emulate.c @@ -261,7 +261,7 @@ static int hvmemul_do_io( * an ioreq server that can handle it. * * Rules: - * A> PIO or MMIO accesses run through hvm_select_ioreq_server() to + * A> PIO or MMIO accesses run through ioreq_server_select() to * choose the ioreq server by range. If no server is found, the access * is ignored. * @@ -323,7 +323,7 @@ static int hvmemul_do_io( } if ( !s ) - s = hvm_select_ioreq_server(currd, &p); + s = ioreq_server_select(currd, &p); /* If there is no suitable backing DM, just ignore accesses */ if ( !s ) @@ -333,7 +333,7 @@ static int hvmemul_do_io( } else { - rc = hvm_send_ioreq(s, &p, 0); + rc = ioreq_send(s, &p, 0); if ( rc != X86EMUL_RETRY || currd->is_shutting_down ) vio->req.state = STATE_IOREQ_NONE; else if ( !ioreq_needs_completion(&vio->req) ) diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c index 4ed929c..0d7bb42 100644 --- a/xen/arch/x86/hvm/hvm.c +++ b/xen/arch/x86/hvm/hvm.c @@ -546,7 +546,7 @@ void hvm_do_resume(struct vcpu *v) pt_restore_timer(v); - if ( !handle_hvm_io_completion(v) ) + if ( !vcpu_ioreq_handle_completion(v) ) return; if ( unlikely(v->arch.vm_event) ) @@ -677,7 +677,7 @@ int hvm_domain_initialise(struct domain *d) register_g2m_portio_handler(d); register_vpci_portio_handler(d); - hvm_ioreq_init(d); + ioreq_domain_init(d); hvm_init_guest_time(d); @@ -739,7 +739,7 @@ void hvm_domain_relinquish_resources(struct domain *d) viridian_domain_deinit(d); - hvm_destroy_all_ioreq_servers(d); + ioreq_server_destroy_all(d); msixtbl_pt_cleanup(d); @@ -1582,7 +1582,7 @@ int hvm_vcpu_initialise(struct vcpu *v) if ( rc ) goto fail5; - rc = hvm_all_ioreq_servers_add_vcpu(d, v); + rc = ioreq_server_add_vcpu_all(d, v); if ( rc != 0 ) goto fail6; @@ -1618,7 +1618,7 @@ void hvm_vcpu_destroy(struct vcpu *v) { viridian_vcpu_deinit(v); - hvm_all_ioreq_servers_remove_vcpu(v->domain, v); + ioreq_server_remove_vcpu_all(v->domain, v); if ( hvm_altp2m_supported() ) altp2m_vcpu_destroy(v); diff --git a/xen/arch/x86/hvm/io.c b/xen/arch/x86/hvm/io.c index dd733e1..66a37ee 100644 --- a/xen/arch/x86/hvm/io.c +++ b/xen/arch/x86/hvm/io.c @@ -60,7 +60,7 @@ void send_timeoffset_req(unsigned long timeoff) if ( timeoff == 0 ) return; - if ( hvm_broadcast_ioreq(&p, true) != 0 ) + if ( ioreq_broadcast(&p, true) != 0 ) gprintk(XENLOG_ERR, "Unsuccessful timeoffset update\n"); } @@ -74,7 +74,7 @@ void send_invalidate_req(void) .data = ~0UL, /* flush all */ }; - if ( hvm_broadcast_ioreq(&p, false) != 0 ) + if ( ioreq_broadcast(&p, false) != 0 ) gprintk(XENLOG_ERR, "Unsuccessful map-cache invalidate\n"); } @@ -155,7 +155,7 @@ bool handle_pio(uint16_t port, unsigned int size, int dir) * We should not advance RIP/EIP if the domain is shutting down or * if X86EMUL_RETRY has been returned by an internal handler. */ - if ( curr->domain->is_shutting_down || !hvm_io_pending(curr) ) + if ( curr->domain->is_shutting_down || !vcpu_ioreq_pending(curr) ) return false; break; diff --git a/xen/arch/x86/hvm/ioreq.c b/xen/arch/x86/hvm/ioreq.c index 62a4b33..02ad9db 100644 --- a/xen/arch/x86/hvm/ioreq.c +++ b/xen/arch/x86/hvm/ioreq.c @@ -153,7 +153,7 @@ static int hvm_map_ioreq_gfn(struct ioreq_server *s, bool buf) { /* * If a page has already been allocated (which will happen on - * demand if hvm_get_ioreq_server_frame() is called), then + * demand if ioreq_server_get_frame() is called), then * mapping a guest frame is not permitted. */ if ( gfn_eq(iorp->gfn, INVALID_GFN) ) diff --git a/xen/arch/x86/hvm/stdvga.c b/xen/arch/x86/hvm/stdvga.c index ee13449..ab9781d 100644 --- a/xen/arch/x86/hvm/stdvga.c +++ b/xen/arch/x86/hvm/stdvga.c @@ -507,11 +507,11 @@ static int stdvga_mem_write(const struct hvm_io_handler *handler, } done: - srv = hvm_select_ioreq_server(current->domain, &p); + srv = ioreq_server_select(current->domain, &p); if ( !srv ) return X86EMUL_UNHANDLEABLE; - return hvm_send_ioreq(srv, &p, 1); + return ioreq_send(srv, &p, 1); } static bool_t stdvga_mem_accept(const struct hvm_io_handler *handler, diff --git a/xen/arch/x86/hvm/vmx/vvmx.c b/xen/arch/x86/hvm/vmx/vvmx.c index 0ddb6a4..e9f94da 100644 --- a/xen/arch/x86/hvm/vmx/vvmx.c +++ b/xen/arch/x86/hvm/vmx/vvmx.c @@ -1517,7 +1517,7 @@ void nvmx_switch_guest(void) * don't want to continue as this setup is not implemented nor supported * as of right now. */ - if ( hvm_io_pending(v) ) + if ( vcpu_ioreq_pending(v) ) return; /* * a softirq may interrupt us between a virtual vmentry is diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index ce3ef59..de3066a 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -59,7 +59,7 @@ static struct ioreq_server *get_ioreq_server(const struct domain *d, * Iterate over all possible ioreq servers. * * NOTE: The iteration is backwards such that more recently created - * ioreq servers are favoured in hvm_select_ioreq_server(). + * ioreq servers are favoured in ioreq_server_select(). * This is a semantic that previously existed when ioreq servers * were held in a linked list. */ @@ -106,12 +106,12 @@ static struct ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, return NULL; } -bool hvm_io_pending(struct vcpu *v) +bool vcpu_ioreq_pending(struct vcpu *v) { return get_pending_vcpu(v, NULL); } -static bool hvm_wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) +static bool wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) { unsigned int prev_state = STATE_IOREQ_NONE; unsigned int state = p->state; @@ -168,7 +168,7 @@ static bool hvm_wait_for_io(struct ioreq_vcpu *sv, ioreq_t *p) return true; } -bool handle_hvm_io_completion(struct vcpu *v) +bool vcpu_ioreq_handle_completion(struct vcpu *v) { struct domain *d = v->domain; struct vcpu_io *vio = &v->io; @@ -183,7 +183,7 @@ bool handle_hvm_io_completion(struct vcpu *v) } sv = get_pending_vcpu(v, &s); - if ( sv && !hvm_wait_for_io(sv, get_ioreq(s, v)) ) + if ( sv && !wait_for_io(sv, get_ioreq(s, v)) ) return false; vio->req.state = ioreq_needs_completion(&vio->req) ? @@ -214,7 +214,7 @@ bool handle_hvm_io_completion(struct vcpu *v) return true; } -static int hvm_alloc_ioreq_mfn(struct ioreq_server *s, bool buf) +static int ioreq_server_alloc_mfn(struct ioreq_server *s, bool buf) { struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; struct page_info *page; @@ -262,7 +262,7 @@ static int hvm_alloc_ioreq_mfn(struct ioreq_server *s, bool buf) return -ENOMEM; } -static void hvm_free_ioreq_mfn(struct ioreq_server *s, bool buf) +static void ioreq_server_free_mfn(struct ioreq_server *s, bool buf) { struct ioreq_page *iorp = buf ? &s->bufioreq : &s->ioreq; struct page_info *page = iorp->page; @@ -301,8 +301,8 @@ bool is_ioreq_server_page(struct domain *d, const struct page_info *page) return found; } -static void hvm_update_ioreq_evtchn(struct ioreq_server *s, - struct ioreq_vcpu *sv) +static void ioreq_server_update_evtchn(struct ioreq_server *s, + struct ioreq_vcpu *sv) { ASSERT(spin_is_locked(&s->lock)); @@ -314,8 +314,8 @@ static void hvm_update_ioreq_evtchn(struct ioreq_server *s, } } -static int hvm_ioreq_server_add_vcpu(struct ioreq_server *s, - struct vcpu *v) +static int ioreq_server_add_vcpu(struct ioreq_server *s, + struct vcpu *v) { struct ioreq_vcpu *sv; int rc; @@ -350,7 +350,7 @@ static int hvm_ioreq_server_add_vcpu(struct ioreq_server *s, list_add(&sv->list_entry, &s->ioreq_vcpu_list); if ( s->enabled ) - hvm_update_ioreq_evtchn(s, sv); + ioreq_server_update_evtchn(s, sv); spin_unlock(&s->lock); return 0; @@ -366,8 +366,8 @@ static int hvm_ioreq_server_add_vcpu(struct ioreq_server *s, return rc; } -static void hvm_ioreq_server_remove_vcpu(struct ioreq_server *s, - struct vcpu *v) +static void ioreq_server_remove_vcpu(struct ioreq_server *s, + struct vcpu *v) { struct ioreq_vcpu *sv; @@ -394,7 +394,7 @@ static void hvm_ioreq_server_remove_vcpu(struct ioreq_server *s, spin_unlock(&s->lock); } -static void hvm_ioreq_server_remove_all_vcpus(struct ioreq_server *s) +static void ioreq_server_remove_all_vcpus(struct ioreq_server *s) { struct ioreq_vcpu *sv, *next; @@ -420,28 +420,28 @@ static void hvm_ioreq_server_remove_all_vcpus(struct ioreq_server *s) spin_unlock(&s->lock); } -static int hvm_ioreq_server_alloc_pages(struct ioreq_server *s) +static int ioreq_server_alloc_pages(struct ioreq_server *s) { int rc; - rc = hvm_alloc_ioreq_mfn(s, false); + rc = ioreq_server_alloc_mfn(s, false); if ( !rc && (s->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) ) - rc = hvm_alloc_ioreq_mfn(s, true); + rc = ioreq_server_alloc_mfn(s, true); if ( rc ) - hvm_free_ioreq_mfn(s, false); + ioreq_server_free_mfn(s, false); return rc; } -static void hvm_ioreq_server_free_pages(struct ioreq_server *s) +static void ioreq_server_free_pages(struct ioreq_server *s) { - hvm_free_ioreq_mfn(s, true); - hvm_free_ioreq_mfn(s, false); + ioreq_server_free_mfn(s, true); + ioreq_server_free_mfn(s, false); } -static void hvm_ioreq_server_free_rangesets(struct ioreq_server *s) +static void ioreq_server_free_rangesets(struct ioreq_server *s) { unsigned int i; @@ -449,8 +449,8 @@ static void hvm_ioreq_server_free_rangesets(struct ioreq_server *s) rangeset_destroy(s->range[i]); } -static int hvm_ioreq_server_alloc_rangesets(struct ioreq_server *s, - ioservid_t id) +static int ioreq_server_alloc_rangesets(struct ioreq_server *s, + ioservid_t id) { unsigned int i; int rc; @@ -482,12 +482,12 @@ static int hvm_ioreq_server_alloc_rangesets(struct ioreq_server *s, return 0; fail: - hvm_ioreq_server_free_rangesets(s); + ioreq_server_free_rangesets(s); return rc; } -static void hvm_ioreq_server_enable(struct ioreq_server *s) +static void ioreq_server_enable(struct ioreq_server *s) { struct ioreq_vcpu *sv; @@ -503,13 +503,13 @@ static void hvm_ioreq_server_enable(struct ioreq_server *s) list_for_each_entry ( sv, &s->ioreq_vcpu_list, list_entry ) - hvm_update_ioreq_evtchn(s, sv); + ioreq_server_update_evtchn(s, sv); done: spin_unlock(&s->lock); } -static void hvm_ioreq_server_disable(struct ioreq_server *s) +static void ioreq_server_disable(struct ioreq_server *s) { spin_lock(&s->lock); @@ -524,9 +524,9 @@ static void hvm_ioreq_server_disable(struct ioreq_server *s) spin_unlock(&s->lock); } -static int hvm_ioreq_server_init(struct ioreq_server *s, - struct domain *d, int bufioreq_handling, - ioservid_t id) +static int ioreq_server_init(struct ioreq_server *s, + struct domain *d, int bufioreq_handling, + ioservid_t id) { struct domain *currd = current->domain; struct vcpu *v; @@ -544,7 +544,7 @@ static int hvm_ioreq_server_init(struct ioreq_server *s, s->ioreq.gfn = INVALID_GFN; s->bufioreq.gfn = INVALID_GFN; - rc = hvm_ioreq_server_alloc_rangesets(s, id); + rc = ioreq_server_alloc_rangesets(s, id); if ( rc ) return rc; @@ -552,7 +552,7 @@ static int hvm_ioreq_server_init(struct ioreq_server *s, for_each_vcpu ( d, v ) { - rc = hvm_ioreq_server_add_vcpu(s, v); + rc = ioreq_server_add_vcpu(s, v); if ( rc ) goto fail_add; } @@ -560,23 +560,23 @@ static int hvm_ioreq_server_init(struct ioreq_server *s, return 0; fail_add: - hvm_ioreq_server_remove_all_vcpus(s); + ioreq_server_remove_all_vcpus(s); arch_ioreq_server_unmap_pages(s); - hvm_ioreq_server_free_rangesets(s); + ioreq_server_free_rangesets(s); put_domain(s->emulator); return rc; } -static void hvm_ioreq_server_deinit(struct ioreq_server *s) +static void ioreq_server_deinit(struct ioreq_server *s) { ASSERT(!s->enabled); - hvm_ioreq_server_remove_all_vcpus(s); + ioreq_server_remove_all_vcpus(s); /* * NOTE: It is safe to call both arch_ioreq_server_unmap_pages() and - * hvm_ioreq_server_free_pages() in that order. + * ioreq_server_free_pages() in that order. * This is because the former will do nothing if the pages * are not mapped, leaving the page to be freed by the latter. * However if the pages are mapped then the former will set @@ -584,9 +584,9 @@ static void hvm_ioreq_server_deinit(struct ioreq_server *s) * nothing. */ arch_ioreq_server_unmap_pages(s); - hvm_ioreq_server_free_pages(s); + ioreq_server_free_pages(s); - hvm_ioreq_server_free_rangesets(s); + ioreq_server_free_rangesets(s); put_domain(s->emulator); } @@ -620,11 +620,11 @@ static int ioreq_server_create(struct domain *d, int bufioreq_handling, /* * It is safe to call set_ioreq_server() prior to - * hvm_ioreq_server_init() since the target domain is paused. + * ioreq_server_init() since the target domain is paused. */ set_ioreq_server(d, i, s); - rc = hvm_ioreq_server_init(s, d, bufioreq_handling, i); + rc = ioreq_server_init(s, d, bufioreq_handling, i); if ( rc ) { set_ioreq_server(d, i, NULL); @@ -668,13 +668,13 @@ static int ioreq_server_destroy(struct domain *d, ioservid_t id) arch_ioreq_server_destroy(s); - hvm_ioreq_server_disable(s); + ioreq_server_disable(s); /* - * It is safe to call hvm_ioreq_server_deinit() prior to + * It is safe to call ioreq_server_deinit() prior to * set_ioreq_server() since the target domain is paused. */ - hvm_ioreq_server_deinit(s); + ioreq_server_deinit(s); set_ioreq_server(d, id, NULL); domain_unpause(d); @@ -736,8 +736,8 @@ static int ioreq_server_get_info(struct domain *d, ioservid_t id, return rc; } -int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, - unsigned long idx, mfn_t *mfn) +int ioreq_server_get_frame(struct domain *d, ioservid_t id, + unsigned long idx, mfn_t *mfn) { struct ioreq_server *s; int rc; @@ -756,7 +756,7 @@ int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, if ( s->emulator != current->domain ) goto out; - rc = hvm_ioreq_server_alloc_pages(s); + rc = ioreq_server_alloc_pages(s); if ( rc ) goto out; @@ -955,9 +955,9 @@ static int ioreq_server_set_state(struct domain *d, ioservid_t id, domain_pause(d); if ( enabled ) - hvm_ioreq_server_enable(s); + ioreq_server_enable(s); else - hvm_ioreq_server_disable(s); + ioreq_server_disable(s); domain_unpause(d); @@ -968,7 +968,7 @@ static int ioreq_server_set_state(struct domain *d, ioservid_t id, return rc; } -int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) +int ioreq_server_add_vcpu_all(struct domain *d, struct vcpu *v) { struct ioreq_server *s; unsigned int id; @@ -978,7 +978,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) FOR_EACH_IOREQ_SERVER(d, id, s) { - rc = hvm_ioreq_server_add_vcpu(s, v); + rc = ioreq_server_add_vcpu(s, v); if ( rc ) goto fail; } @@ -995,7 +995,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) if ( !s ) continue; - hvm_ioreq_server_remove_vcpu(s, v); + ioreq_server_remove_vcpu(s, v); } spin_unlock_recursive(&d->ioreq_server.lock); @@ -1003,7 +1003,7 @@ int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v) return rc; } -void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) +void ioreq_server_remove_vcpu_all(struct domain *d, struct vcpu *v) { struct ioreq_server *s; unsigned int id; @@ -1011,12 +1011,12 @@ void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v) spin_lock_recursive(&d->ioreq_server.lock); FOR_EACH_IOREQ_SERVER(d, id, s) - hvm_ioreq_server_remove_vcpu(s, v); + ioreq_server_remove_vcpu(s, v); spin_unlock_recursive(&d->ioreq_server.lock); } -void hvm_destroy_all_ioreq_servers(struct domain *d) +void ioreq_server_destroy_all(struct domain *d) { struct ioreq_server *s; unsigned int id; @@ -1030,13 +1030,13 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) FOR_EACH_IOREQ_SERVER(d, id, s) { - hvm_ioreq_server_disable(s); + ioreq_server_disable(s); /* - * It is safe to call hvm_ioreq_server_deinit() prior to + * It is safe to call ioreq_server_deinit() prior to * set_ioreq_server() since the target domain is being destroyed. */ - hvm_ioreq_server_deinit(s); + ioreq_server_deinit(s); set_ioreq_server(d, id, NULL); xfree(s); @@ -1045,8 +1045,8 @@ void hvm_destroy_all_ioreq_servers(struct domain *d) spin_unlock_recursive(&d->ioreq_server.lock); } -struct ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p) +struct ioreq_server *ioreq_server_select(struct domain *d, + ioreq_t *p) { struct ioreq_server *s; uint8_t type; @@ -1101,7 +1101,7 @@ struct ioreq_server *hvm_select_ioreq_server(struct domain *d, return NULL; } -static int hvm_send_buffered_ioreq(struct ioreq_server *s, ioreq_t *p) +static int ioreq_send_buffered(struct ioreq_server *s, ioreq_t *p) { struct domain *d = current->domain; struct ioreq_page *iorp; @@ -1194,8 +1194,8 @@ static int hvm_send_buffered_ioreq(struct ioreq_server *s, ioreq_t *p) return IOREQ_STATUS_HANDLED; } -int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, - bool buffered) +int ioreq_send(struct ioreq_server *s, ioreq_t *proto_p, + bool buffered) { struct vcpu *curr = current; struct domain *d = curr->domain; @@ -1204,7 +1204,7 @@ int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, ASSERT(s); if ( buffered ) - return hvm_send_buffered_ioreq(s, proto_p); + return ioreq_send_buffered(s, proto_p); if ( unlikely(!vcpu_start_shutdown_deferral(curr)) ) return IOREQ_STATUS_RETRY; @@ -1254,7 +1254,7 @@ int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, return IOREQ_STATUS_UNHANDLED; } -unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) +unsigned int ioreq_broadcast(ioreq_t *p, bool buffered) { struct domain *d = current->domain; struct ioreq_server *s; @@ -1265,14 +1265,14 @@ unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered) if ( !s->enabled ) continue; - if ( hvm_send_ioreq(s, p, buffered) == IOREQ_STATUS_UNHANDLED ) + if ( ioreq_send(s, p, buffered) == IOREQ_STATUS_UNHANDLED ) failed++; } return failed; } -void hvm_ioreq_init(struct domain *d) +void ioreq_domain_init(struct domain *d) { spin_lock_init(&d->ioreq_server.lock); diff --git a/xen/common/memory.c b/xen/common/memory.c index 2f274a6..8eb05b1 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -1113,7 +1113,7 @@ static int acquire_ioreq_server(struct domain *d, { mfn_t mfn; - rc = hvm_get_ioreq_server_frame(d, id, frame + i, &mfn); + rc = ioreq_server_get_frame(d, id, frame + i, &mfn); if ( rc ) return rc; diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index eace1d3..0b433e2 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -83,26 +83,26 @@ static inline bool ioreq_needs_completion(const ioreq_t *ioreq) #define HANDLE_BUFIOREQ(s) \ ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) -bool hvm_io_pending(struct vcpu *v); -bool handle_hvm_io_completion(struct vcpu *v); +bool vcpu_ioreq_pending(struct vcpu *v); +bool vcpu_ioreq_handle_completion(struct vcpu *v); bool is_ioreq_server_page(struct domain *d, const struct page_info *page); -int hvm_get_ioreq_server_frame(struct domain *d, ioservid_t id, - unsigned long idx, mfn_t *mfn); +int ioreq_server_get_frame(struct domain *d, ioservid_t id, + unsigned long idx, mfn_t *mfn); int ioreq_server_map_mem_type(struct domain *d, ioservid_t id, uint32_t type, uint32_t flags); -int hvm_all_ioreq_servers_add_vcpu(struct domain *d, struct vcpu *v); -void hvm_all_ioreq_servers_remove_vcpu(struct domain *d, struct vcpu *v); -void hvm_destroy_all_ioreq_servers(struct domain *d); +int ioreq_server_add_vcpu_all(struct domain *d, struct vcpu *v); +void ioreq_server_remove_vcpu_all(struct domain *d, struct vcpu *v); +void ioreq_server_destroy_all(struct domain *d); -struct ioreq_server *hvm_select_ioreq_server(struct domain *d, - ioreq_t *p); -int hvm_send_ioreq(struct ioreq_server *s, ioreq_t *proto_p, - bool buffered); -unsigned int hvm_broadcast_ioreq(ioreq_t *p, bool buffered); +struct ioreq_server *ioreq_server_select(struct domain *d, + ioreq_t *p); +int ioreq_send(struct ioreq_server *s, ioreq_t *proto_p, + bool buffered); +unsigned int ioreq_broadcast(ioreq_t *p, bool buffered); -void hvm_ioreq_init(struct domain *d); +void ioreq_domain_init(struct domain *d); int ioreq_server_dm_op(struct xen_dm_op *op, struct domain *d, bool *const_op); From patchwork Mon Jan 25 19:08:20 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044077 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5586FC433E0 for ; Mon, 25 Jan 2021 19:19:16 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id DCEE520E65 for ; Mon, 25 Jan 2021 19:19:15 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org DCEE520E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74337.133609 (Exim 4.92) (envelope-from ) id 1l47O4-0001q8-Ln; Mon, 25 Jan 2021 19:18:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74337.133609; Mon, 25 Jan 2021 19:18:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47O4-0001q1-IA; Mon, 25 Jan 2021 19:18:56 +0000 Received: by outflank-mailman (input) for mailman id 74337; Mon, 25 Jan 2021 19:18:55 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47FV-0008N7-Ho for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:05 +0000 Received: from mail-wm1-x32f.google.com (unknown [2a00:1450:4864:20::32f]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 79271237-14c8-4b31-ad35-d19f1539bb20; Mon, 25 Jan 2021 19:09:07 +0000 (UTC) Received: by mail-wm1-x32f.google.com with SMTP id f16so366644wmq.5 for ; Mon, 25 Jan 2021 11:09:07 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.06 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:06 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 79271237-14c8-4b31-ad35-d19f1539bb20 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=VAsD+FeaZvy6J0TsxqgptDfccADq1PaOx6WnTwbiR+g=; b=Dlh9p5G57YHlcrwU8i/TTCx4reu0wzKDXrx4OabJDzpPyCEf95g2GRexZQUoLZXmRE Uo+gNL77F4QItzjb/V2BbJu4LVn/52dlal+b5rNxbnNt6cTPSjOE5SRPmCVYQHMpGPDL UKtOjCFwmnIF1iUoSkVNYCMtpI158FtB003xjCyOjcM6AoXShET2cr6VQOyfwI6c5lcK 7L4yKvYo6oFIAER3i93FC0shfV5nfg0jrd1gQcfiJAj+eaaCbCJ18MYLnfRww7hLQjTm RhCdNUYbPuM5BZG5rDe4ZO7jmcjheJBPfOmIS0PpQlixIf6pAdSHYFFhnXXm2Hr6LD8R 4iqQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=VAsD+FeaZvy6J0TsxqgptDfccADq1PaOx6WnTwbiR+g=; b=LIqYFnyXZGelv5anluWg+TAgzZfNNIQvoHSnD5OqtxLIbuX2rY3vFxFo+yX1IaLwJH MAJfLX1eh/bdACG3XYjSlwFZyFpx3AF77M+AmnJIvfhT9eW5hFnCJ9abzXK66v1nnPZT 9m6h/I0iUS/bioglE7d6W4YlRKyl1PGfvePFQU3bt1KeZ4ttOuJkSts0K/bLHkTl5oEG 6SYM6zA970YuvPKUtzB3jT97igkq7zIX/5wdQrivSoVAOPrDscPa5HrdoVCL3cy5yJpl xc8QqHmC0eYNY9B1xyfIJ2A4wTI+ditCEv5Iqp4QaPCtcb8bctyqtw0maJGjy7nueu0F 6SnQ== X-Gm-Message-State: AOAM5311R2k2Ay87+vXGC2hFv9IWpxlk6W/3v5dg/kzpT2Dz18qHaJBk 8OaOU9JUIMQ+NvW2KF+9Q4l5elglahEf9g== X-Google-Smtp-Source: ABdhPJxeEkAcuxneuHRgiPv7tP1j0g/ocU+Y0qchoFbVU7tSJ2sYYSGBuMxfBTx2ZyYurMqAGzsyOA== X-Received: by 2002:a1c:2d8a:: with SMTP id t132mr1391415wmt.119.1611601746819; Mon, 25 Jan 2021 11:09:06 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Paul Durrant , Stefano Stabellini , Julien Grall , Julien Grall Subject: [PATCH V5 13/22] xen/ioreq: Use guest_cmpxchg64() instead of cmpxchg() Date: Mon, 25 Jan 2021 21:08:20 +0200 Message-Id: <1611601709-28361-14-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko The cmpxchg() in ioreq_send_buffered() operates on memory shared with the emulator domain (and the target domain if the legacy interface is used). In order to be on the safe side we need to switch to guest_cmpxchg64() to prevent a domain to DoS Xen on Arm. The point to use 64-bit version of helper is to support Arm32 since the IOREQ code uses cmpxchg() with 64-bit value. As there is no plan to support the legacy interface on Arm, we will have a page to be mapped in a single domain at the time, so we can use s->emulator in guest_cmpxchg64() safely. Thankfully the only user of the legacy interface is x86 so far and there is not concern regarding the atomics operations. Please note, that the legacy interface *must* not be used on Arm without revisiting the code. Signed-off-by: Oleksandr Tyshchenko Acked-by: Stefano Stabellini Reviewed-by: Paul Durrant CC: Julien Grall [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch Changes V1 -> V2: - move earlier to avoid breaking arm32 compilation - add an explanation to commit description and hvm_allow_set_param() - pass s->emulator Changes V2 -> V3: - update patch description Changes V3 -> V4: - add Stefano's A-b - drop comment from arm/hvm.c Changes V4 -> V5: - update patch description - rebase - add Paul's R-b --- --- xen/common/ioreq.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index de3066a..07572a5 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -29,6 +29,7 @@ #include #include +#include #include #include @@ -1185,7 +1186,7 @@ static int ioreq_send_buffered(struct ioreq_server *s, ioreq_t *p) new.read_pointer = old.read_pointer - n * IOREQ_BUFFER_SLOT_NUM; new.write_pointer = old.write_pointer - n * IOREQ_BUFFER_SLOT_NUM; - cmpxchg(&pg->ptrs.full, old.full, new.full); + guest_cmpxchg64(s->emulator, &pg->ptrs.full, old.full, new.full); } notify_via_xen_event_channel(d, s->bufioreq_evtchn); From patchwork Mon Jan 25 19:08:21 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044103 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 961EEC433DB for ; Mon, 25 Jan 2021 19:21:20 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 2F2E12251F for ; Mon, 25 Jan 2021 19:21:20 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 2F2E12251F Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74380.133735 (Exim 4.92) (envelope-from ) id 1l47QC-0003f3-HO; Mon, 25 Jan 2021 19:21:08 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74380.133735; Mon, 25 Jan 2021 19:21:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47QC-0003eo-AT; Mon, 25 Jan 2021 19:21:08 +0000 Received: by outflank-mailman (input) for mailman id 74380; Mon, 25 Jan 2021 19:21:06 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Fa-0008N7-Hw for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:10 +0000 Received: from mail-wm1-x32a.google.com (unknown [2a00:1450:4864:20::32a]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 59038c9e-4d4c-4b91-93bd-68045fd04504; Mon, 25 Jan 2021 19:09:09 +0000 (UTC) Received: by mail-wm1-x32a.google.com with SMTP id m2so398846wmm.1 for ; Mon, 25 Jan 2021 11:09:09 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.06 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:07 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 59038c9e-4d4c-4b91-93bd-68045fd04504 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=spVL9biPBCQHdDaHpN1ip3vXv4EE+lr+qlUtmYfQbBM=; b=D8zcFMRc7drV9zaSaKc+QuPopcsJ9siS4sOCfLvFXh6+Yflp0HUzP4FZybaW19n1Y7 U1h+t7j1ecYHHyNOkXpD10613/xqrfA+hjtxAb/rdiz8nst5wVEAyfBjvXZgX+cg4B6z 7CT+Cp59iAOs27GS1HIabNFafbxl0NjfzQ2tHhaN5ODrUFWygU4N4BSSIq/F8aWJ44Ub rdtkQ76kvdVx3nk4wxso7nttQLNCEDG2DVQEdSiCfWlgow/qN6+OC8KfH7KLAqIRe0/Q nFC5EiEQFMMbFyt3YKJR0N5g5wUmz5j6Frfd4wor3pzbqVWnFv/Q3mmwRFb2cYSuMwbu TG3g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=spVL9biPBCQHdDaHpN1ip3vXv4EE+lr+qlUtmYfQbBM=; b=IoRb0VgND0beAqerFPlfyUVlUABsULvWktT6BtjqpP03InCP0s7PbYBVjZ1WoY9VfW HZ+dK/Uxe2oGgaumjRklq7URIj5f2QNf//OHBQo89UspOZSyevROT6WDcjNHkqnT78k1 l1DiJ7N7ufE8wcWfkIngdhEqjuXeut/88oDq0L22oeydKm/r93eomVcjtcA089x/UMAM MWgkMthdtJn9lPQiSi/7wG/inhja4FomheKvmGpHgMxwbNwVz8CCxT1k8GNfT2Cu1FeY zGPOFiFejkXRzE8/c7Tnk5OJ473lNQQhxoaskHmG9vSAYnZN1uf+JXkFNLWDBTbCy4aG HhJQ== X-Gm-Message-State: AOAM531fHXLVMagSjr+QLv3k/MjMlmRD4QYq1otvSu+srgn7HX1Pz2UN Z/z9GNtBevhV7LD2q/0ttsVmBZW5ES+T3g== X-Google-Smtp-Source: ABdhPJy1FZ8i5nLl3Soumms/fQfX0NWzlhZxSPgc1k/YrUtw6jAUUyAhKFKOy/nSxLMvWsC89nnmzQ== X-Received: by 2002:a1c:1f86:: with SMTP id f128mr1442919wmf.174.1611601747969; Mon, 25 Jan 2021 11:09:07 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Julien Grall , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Andrew Cooper , George Dunlap , Ian Jackson , Jan Beulich , Wei Liu , Oleksandr Tyshchenko Subject: [PATCH V5 14/22] arm/ioreq: Introduce arch specific bits for IOREQ/DM features Date: Mon, 25 Jan 2021 21:08:21 +0200 Message-Id: <1611601709-28361-15-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Julien Grall This patch adds basic IOREQ/DM support on Arm. The subsequent patches will improve functionality and add remaining bits. The IOREQ/DM features are supposed to be built with IOREQ_SERVER option enabled, which is disabled by default on Arm for now. Please note, the "PIO handling" TODO is expected to left unaddressed for the current series. It is not an big issue for now while Xen doesn't have support for vPCI on Arm. On Arm64 they are only used for PCI IO Bar and we would probably want to expose them to emulator as PIO access to make a DM completely arch-agnostic. So "PIO handling" should be implemented when we add support for vPCI. Signed-off-by: Julien Grall Signed-off-by: Oleksandr Tyshchenko [On Arm only] Tested-by: Wei Chen Reviewed-by: Stefano Stabellini --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" *** I admit, I didn't resolve header dependencies completely. For now, public/hvm/dm_op.h is included by xen/dm.h, but ought to be included by arch/arm/dm.c. Details here: https://lore.kernel.org/xen-devel/e0bc7f80-974e-945d-4605-173bd05302af@gmail.com/ *** Changes RFC -> V1: - was split into: - arm/ioreq: Introduce arch specific bits for IOREQ/DM features - xen/mm: Handle properly reference in set_foreign_p2m_entry() on Arm - update patch description - update asm-arm/hvm/ioreq.h according to the newly introduced arch functions: - arch_hvm_destroy_ioreq_server() - arch_handle_hvm_io_completion() - update arch files to include xen/ioreq.h - remove HVMOP plumbing - rewrite a logic to handle properly case when hvm_send_ioreq() returns IO_RETRY - add a logic to handle properly handle_hvm_io_completion() return value - rename handle_mmio() to ioreq_handle_complete_mmio() - move paging_mark_pfn_dirty() to asm-arm/paging.h - remove forward declaration for hvm_ioreq_server in asm-arm/paging.h - move try_fwd_ioserv() to ioreq.c, provide stubs if !CONFIG_IOREQ_SERVER - do not remove #ifdef CONFIG_IOREQ_SERVER in memory.c for guarding xen/ioreq.h - use gdprintk in try_fwd_ioserv(), remove unneeded prints - update list of #include-s - move has_vpci() to asm-arm/domain.h - add a comment (TODO) to unimplemented yet handle_pio() - remove hvm_mmio_first(last)_byte() and hvm_ioreq_(page/vcpu/server) structs from the arch files, they were already moved to the common code - remove set_foreign_p2m_entry() changes, they will be properly implemented in the follow-up patch - select IOREQ_SERVER for Arm instead of Arm64 in Kconfig - remove x86's realmode and other unneeded stubs from xen/ioreq.h - clafify ioreq_t p.df usage in try_fwd_ioserv() - set ioreq_t p.count to 1 in try_fwd_ioserv() Changes V1 -> V2: - was split into: - arm/ioreq: Introduce arch specific bits for IOREQ/DM features - xen/arm: Stick around in leave_hypervisor_to_guest until I/O has completed - update the author of a patch - update patch description - move a loop in leave_hypervisor_to_guest() to a separate patch - set IOREQ_SERVER disabled by default - remove already clarified /* XXX */ - replace BUG() by ASSERT_UNREACHABLE() in handle_pio() - remove default case for handling the return value of try_handle_mmio() - remove struct hvm_domain, enum hvm_io_completion, struct hvm_vcpu_io, struct hvm_vcpu from asm-arm/domain.h, these are common materials now - update everything according to the recent changes (IOREQ related function names don't contain "hvm" prefixes/infixes anymore, IOREQ related fields are part of common struct vcpu/domain now, etc) Changes V2 -> V3: - update patch according the "legacy interface" is x86 specific - add dummy arch hooks - remove dummy paging_mark_pfn_dirty() - don’t include in common ioreq.c - don’t include in arch ioreq.h - remove #define ioreq_params(d, i) Changes V3 -> V4: - rebase - update patch according to the renaming IO_ -> VIO_ (io_ -> vio_) and misc changes to arch hooks - update patch according to the IOREQ related dm-op handling changes - don't include from arch header - make all arch hooks out-of-line - add a comment above IOREQ_STATUS_* #define-s Changes V4 -> V5: - change the placement of ioreq_server_destroy_all() in arm/domain.c - don't include public/hvm/dm_op.h by asm-arm/domain.h - include public/hvm/dm_op.h by xen/dm.h - put arch ioreq.h directly into asm-arm subdir - remove do_dm_op() in arm/dm.c, this is a common material now - remove obsolete ioreq_complete_mmio() from asm-arm/ioreq.h - optimize arch_ioreq_complete_mmio() to not call try_handle_mmio(), but try_handle_mmio(), use ASSERT_UNREACHABLE() if state is incorrect - split changes to check_for_vcpu_work() to be squashed with patch #15 --- --- xen/arch/arm/Makefile | 2 + xen/arch/arm/dm.c | 97 +++++++++++++++++++ xen/arch/arm/domain.c | 9 ++ xen/arch/arm/io.c | 12 ++- xen/arch/arm/ioreq.c | 215 +++++++++++++++++++++++++++++++++++++++++++ xen/arch/arm/traps.c | 6 ++ xen/include/asm-arm/domain.h | 2 + xen/include/asm-arm/ioreq.h | 70 ++++++++++++++ xen/include/asm-arm/mmio.h | 1 + xen/include/xen/dm.h | 2 + 10 files changed, 415 insertions(+), 1 deletion(-) create mode 100644 xen/arch/arm/dm.c create mode 100644 xen/arch/arm/ioreq.c create mode 100644 xen/include/asm-arm/ioreq.h diff --git a/xen/arch/arm/Makefile b/xen/arch/arm/Makefile index 512ffdd..16e6523 100644 --- a/xen/arch/arm/Makefile +++ b/xen/arch/arm/Makefile @@ -13,6 +13,7 @@ obj-y += cpuerrata.o obj-y += cpufeature.o obj-y += decode.o obj-y += device.o +obj-$(CONFIG_IOREQ_SERVER) += dm.o obj-y += domain.o obj-y += domain_build.init.o obj-y += domctl.o @@ -27,6 +28,7 @@ obj-y += guest_atomics.o obj-y += guest_walk.o obj-y += hvm.o obj-y += io.o +obj-$(CONFIG_IOREQ_SERVER) += ioreq.o obj-y += irq.o obj-y += kernel.init.o obj-$(CONFIG_LIVEPATCH) += livepatch.o diff --git a/xen/arch/arm/dm.c b/xen/arch/arm/dm.c new file mode 100644 index 0000000..f254ed7 --- /dev/null +++ b/xen/arch/arm/dm.c @@ -0,0 +1,97 @@ +/* + * Copyright (c) 2019 Arm ltd. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#include +#include +#include +#include +#include + +int dm_op(const struct dmop_args *op_args) +{ + struct domain *d; + struct xen_dm_op op; + bool const_op = true; + long rc; + size_t offset; + + static const uint8_t op_size[] = { + [XEN_DMOP_create_ioreq_server] = sizeof(struct xen_dm_op_create_ioreq_server), + [XEN_DMOP_get_ioreq_server_info] = sizeof(struct xen_dm_op_get_ioreq_server_info), + [XEN_DMOP_map_io_range_to_ioreq_server] = sizeof(struct xen_dm_op_ioreq_server_range), + [XEN_DMOP_unmap_io_range_from_ioreq_server] = sizeof(struct xen_dm_op_ioreq_server_range), + [XEN_DMOP_set_ioreq_server_state] = sizeof(struct xen_dm_op_set_ioreq_server_state), + [XEN_DMOP_destroy_ioreq_server] = sizeof(struct xen_dm_op_destroy_ioreq_server), + }; + + rc = rcu_lock_remote_domain_by_id(op_args->domid, &d); + if ( rc ) + return rc; + + rc = xsm_dm_op(XSM_DM_PRIV, d); + if ( rc ) + goto out; + + offset = offsetof(struct xen_dm_op, u); + + rc = -EFAULT; + if ( op_args->buf[0].size < offset ) + goto out; + + if ( copy_from_guest_offset((void *)&op, op_args->buf[0].h, 0, offset) ) + goto out; + + if ( op.op >= ARRAY_SIZE(op_size) ) + { + rc = -EOPNOTSUPP; + goto out; + } + + op.op = array_index_nospec(op.op, ARRAY_SIZE(op_size)); + + if ( op_args->buf[0].size < offset + op_size[op.op] ) + goto out; + + if ( copy_from_guest_offset((void *)&op.u, op_args->buf[0].h, offset, + op_size[op.op]) ) + goto out; + + rc = -EINVAL; + if ( op.pad ) + goto out; + + rc = ioreq_server_dm_op(&op, d, &const_op); + + if ( (!rc || rc == -ERESTART) && + !const_op && copy_to_guest_offset(op_args->buf[0].h, offset, + (void *)&op.u, op_size[op.op]) ) + rc = -EFAULT; + + out: + rcu_unlock_domain(d); + + return rc; +} + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/arch/arm/domain.c b/xen/arch/arm/domain.c index 18cafcd..bdd3d3e 100644 --- a/xen/arch/arm/domain.c +++ b/xen/arch/arm/domain.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include #include @@ -696,6 +697,10 @@ int arch_domain_create(struct domain *d, ASSERT(config != NULL); +#ifdef CONFIG_IOREQ_SERVER + ioreq_domain_init(d); +#endif + /* p2m_init relies on some value initialized by the IOMMU subsystem */ if ( (rc = iommu_domain_init(d, config->iommu_opts)) != 0 ) goto fail; @@ -1009,6 +1014,10 @@ int domain_relinquish_resources(struct domain *d) */ domain_vpl011_deinit(d); +#ifdef CONFIG_IOREQ_SERVER + ioreq_server_destroy_all(d); +#endif + PROGRESS(tee): ret = tee_relinquish_resources(d); if (ret ) diff --git a/xen/arch/arm/io.c b/xen/arch/arm/io.c index ae7ef96..7ac0303 100644 --- a/xen/arch/arm/io.c +++ b/xen/arch/arm/io.c @@ -16,12 +16,14 @@ * GNU General Public License for more details. */ +#include #include #include #include #include #include #include +#include #include #include "decode.h" @@ -123,7 +125,15 @@ enum io_state try_handle_mmio(struct cpu_user_regs *regs, handler = find_mmio_handler(v->domain, info.gpa); if ( !handler ) - return IO_UNHANDLED; + { + int rc; + + rc = try_fwd_ioserv(regs, v, &info); + if ( rc == IO_HANDLED ) + return handle_ioserv(regs, v); + + return rc; + } /* All the instructions used on emulated MMIO region should be valid */ if ( !dabt.valid ) diff --git a/xen/arch/arm/ioreq.c b/xen/arch/arm/ioreq.c new file mode 100644 index 0000000..ffeeb0b --- /dev/null +++ b/xen/arch/arm/ioreq.c @@ -0,0 +1,215 @@ +/* + * arm/ioreq.c: hardware virtual machine I/O emulation + * + * Copyright (c) 2019 Arm ltd. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#include +#include + +#include + +#include + +enum io_state handle_ioserv(struct cpu_user_regs *regs, struct vcpu *v) +{ + const union hsr hsr = { .bits = regs->hsr }; + const struct hsr_dabt dabt = hsr.dabt; + /* Code is similar to handle_read */ + uint8_t size = (1 << dabt.size) * 8; + register_t r = v->io.req.data; + + /* We are done with the IO */ + v->io.req.state = STATE_IOREQ_NONE; + + if ( dabt.write ) + return IO_HANDLED; + + /* + * Sign extend if required. + * Note that we expect the read handler to have zeroed the bits + * outside the requested access size. + */ + if ( dabt.sign && (r & (1UL << (size - 1))) ) + { + /* + * We are relying on register_t using the same as + * an unsigned long in order to keep the 32-bit assembly + * code smaller. + */ + BUILD_BUG_ON(sizeof(register_t) != sizeof(unsigned long)); + r |= (~0UL) << size; + } + + set_user_reg(regs, dabt.reg, r); + + return IO_HANDLED; +} + +enum io_state try_fwd_ioserv(struct cpu_user_regs *regs, + struct vcpu *v, mmio_info_t *info) +{ + struct vcpu_io *vio = &v->io; + ioreq_t p = { + .type = IOREQ_TYPE_COPY, + .addr = info->gpa, + .size = 1 << info->dabt.size, + .count = 1, + .dir = !info->dabt.write, + /* + * On x86, df is used by 'rep' instruction to tell the direction + * to iterate (forward or backward). + * On Arm, all the accesses to MMIO region will do a single + * memory access. So for now, we can safely always set to 0. + */ + .df = 0, + .data = get_user_reg(regs, info->dabt.reg), + .state = STATE_IOREQ_READY, + }; + struct ioreq_server *s = NULL; + enum io_state rc; + + switch ( vio->req.state ) + { + case STATE_IOREQ_NONE: + break; + + default: + gdprintk(XENLOG_ERR, "wrong state %u\n", vio->req.state); + return IO_ABORT; + } + + s = ioreq_server_select(v->domain, &p); + if ( !s ) + return IO_UNHANDLED; + + if ( !info->dabt.valid ) + return IO_ABORT; + + vio->req = p; + + rc = ioreq_send(s, &p, 0); + if ( rc != IO_RETRY || v->domain->is_shutting_down ) + vio->req.state = STATE_IOREQ_NONE; + else if ( !ioreq_needs_completion(&vio->req) ) + rc = IO_HANDLED; + else + vio->completion = VIO_mmio_completion; + + return rc; +} + +bool arch_ioreq_complete_mmio(void) +{ + struct vcpu *v = current; + struct cpu_user_regs *regs = guest_cpu_user_regs(); + const union hsr hsr = { .bits = regs->hsr }; + + if ( v->io.req.state != STATE_IORESP_READY ) + { + ASSERT_UNREACHABLE(); + return false; + } + + if ( handle_ioserv(regs, v) == IO_HANDLED ) + { + advance_pc(regs, hsr); + return true; + } + + return false; +} + +bool arch_vcpu_ioreq_completion(enum vio_completion completion) +{ + ASSERT_UNREACHABLE(); + return true; +} + +/* + * The "legacy" mechanism of mapping magic pages for the IOREQ servers + * is x86 specific, so the following hooks don't need to be implemented on Arm: + * - arch_ioreq_server_map_pages + * - arch_ioreq_server_unmap_pages + * - arch_ioreq_server_enable + * - arch_ioreq_server_disable + */ +int arch_ioreq_server_map_pages(struct ioreq_server *s) +{ + return -EOPNOTSUPP; +} + +void arch_ioreq_server_unmap_pages(struct ioreq_server *s) +{ +} + +void arch_ioreq_server_enable(struct ioreq_server *s) +{ +} + +void arch_ioreq_server_disable(struct ioreq_server *s) +{ +} + +void arch_ioreq_server_destroy(struct ioreq_server *s) +{ +} + +int arch_ioreq_server_map_mem_type(struct domain *d, + struct ioreq_server *s, + uint32_t flags) +{ + return -EOPNOTSUPP; +} + +void arch_ioreq_server_map_mem_type_completed(struct domain *d, + struct ioreq_server *s, + uint32_t flags) +{ +} + +bool arch_ioreq_server_destroy_all(struct domain *d) +{ + return true; +} + +bool arch_ioreq_server_get_type_addr(const struct domain *d, + const ioreq_t *p, + uint8_t *type, + uint64_t *addr) +{ + if ( p->type != IOREQ_TYPE_COPY && p->type != IOREQ_TYPE_PIO ) + return false; + + *type = (p->type == IOREQ_TYPE_PIO) ? + XEN_DMOP_IO_RANGE_PORT : XEN_DMOP_IO_RANGE_MEMORY; + *addr = p->addr; + + return true; +} + +void arch_ioreq_domain_init(struct domain *d) +{ +} + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c index 1af1bb9..b0cd8f9 100644 --- a/xen/arch/arm/traps.c +++ b/xen/arch/arm/traps.c @@ -1385,6 +1385,9 @@ static arm_hypercall_t arm_hypercall_table[] = { #ifdef CONFIG_HYPFS HYPERCALL(hypfs_op, 5), #endif +#ifdef CONFIG_IOREQ_SERVER + HYPERCALL(dm_op, 3), +#endif }; #ifndef NDEBUG @@ -1956,6 +1959,9 @@ static void do_trap_stage2_abort_guest(struct cpu_user_regs *regs, case IO_HANDLED: advance_pc(regs, hsr); return; + case IO_RETRY: + /* finish later */ + return; case IO_UNHANDLED: /* IO unhandled, try another way to handle it. */ break; diff --git a/xen/include/asm-arm/domain.h b/xen/include/asm-arm/domain.h index 6819a3b..1da90f2 100644 --- a/xen/include/asm-arm/domain.h +++ b/xen/include/asm-arm/domain.h @@ -262,6 +262,8 @@ static inline void arch_vcpu_block(struct vcpu *v) {} #define arch_vm_assist_valid_mask(d) (1UL << VMASST_TYPE_runstate_update_flag) +#define has_vpci(d) ({ (void)(d); false; }) + #endif /* __ASM_DOMAIN_H__ */ /* diff --git a/xen/include/asm-arm/ioreq.h b/xen/include/asm-arm/ioreq.h new file mode 100644 index 0000000..5018597 --- /dev/null +++ b/xen/include/asm-arm/ioreq.h @@ -0,0 +1,70 @@ +/* + * ioreq.h: Hardware virtual machine assist interface definitions. + * + * Copyright (c) 2016 Citrix Systems Inc. + * Copyright (c) 2019 Arm ltd. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; If not, see . + */ + +#ifndef __ASM_ARM_IOREQ_H__ +#define __ASM_ARM_IOREQ_H__ + +#ifdef CONFIG_IOREQ_SERVER +enum io_state handle_ioserv(struct cpu_user_regs *regs, struct vcpu *v); +enum io_state try_fwd_ioserv(struct cpu_user_regs *regs, + struct vcpu *v, mmio_info_t *info); +#else +static inline enum io_state handle_ioserv(struct cpu_user_regs *regs, + struct vcpu *v) +{ + return IO_UNHANDLED; +} + +static inline enum io_state try_fwd_ioserv(struct cpu_user_regs *regs, + struct vcpu *v, mmio_info_t *info) +{ + return IO_UNHANDLED; +} +#endif + +static inline bool handle_pio(uint16_t port, unsigned int size, int dir) +{ + /* + * TODO: For Arm64, the main user will be PCI. So this should be + * implemented when we add support for vPCI. + */ + ASSERT_UNREACHABLE(); + return true; +} + +static inline void msix_write_completion(struct vcpu *v) +{ +} + +/* This correlation must not be altered */ +#define IOREQ_STATUS_HANDLED IO_HANDLED +#define IOREQ_STATUS_UNHANDLED IO_UNHANDLED +#define IOREQ_STATUS_RETRY IO_RETRY + +#endif /* __ASM_ARM_IOREQ_H__ */ + +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * tab-width: 4 + * indent-tabs-mode: nil + * End: + */ diff --git a/xen/include/asm-arm/mmio.h b/xen/include/asm-arm/mmio.h index 8dbfb27..7ab873c 100644 --- a/xen/include/asm-arm/mmio.h +++ b/xen/include/asm-arm/mmio.h @@ -37,6 +37,7 @@ enum io_state IO_ABORT, /* The IO was handled by the helper and led to an abort. */ IO_HANDLED, /* The IO was successfully handled by the helper. */ IO_UNHANDLED, /* The IO was not handled by the helper. */ + IO_RETRY, /* Retry the emulation for some reason */ }; typedef int (*mmio_read_t)(struct vcpu *v, mmio_info_t *info, diff --git a/xen/include/xen/dm.h b/xen/include/xen/dm.h index 4bd7e57..7f9f0bd 100644 --- a/xen/include/xen/dm.h +++ b/xen/include/xen/dm.h @@ -19,6 +19,8 @@ #include +#include + struct dmop_args { domid_t domid; unsigned int nr_bufs; From patchwork Mon Jan 25 19:08:22 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044101 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id B529EC433E6 for ; Mon, 25 Jan 2021 19:21:18 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 5EEC621D94 for ; Mon, 25 Jan 2021 19:21:18 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 5EEC621D94 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74382.133758 (Exim 4.92) (envelope-from ) id 1l47QD-0003hz-PX; Mon, 25 Jan 2021 19:21:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74382.133758; Mon, 25 Jan 2021 19:21:09 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47QD-0003gq-8c; Mon, 25 Jan 2021 19:21:09 +0000 Received: by outflank-mailman (input) for mailman id 74382; Mon, 25 Jan 2021 19:21:07 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Ff-0008N7-IA for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:15 +0000 Received: from mail-wm1-x32f.google.com (unknown [2a00:1450:4864:20::32f]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id b7276d17-c289-41d0-a450-c6fffe7ed54f; Mon, 25 Jan 2021 19:09:10 +0000 (UTC) Received: by mail-wm1-x32f.google.com with SMTP id c128so545635wme.2 for ; Mon, 25 Jan 2021 11:09:10 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.08 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:08 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: b7276d17-c289-41d0-a450-c6fffe7ed54f DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=3/Dl5PXNfiRS8xSHTBcU7A9OZpqtcxgTL6VyLjhU22w=; b=puuNf0z2jXmPBrNbi98ylRp10BAglq7qS+bINVPDzUxA/cEzPf1RyP0hujQRYj0g8X 6OmrWppphYAhCNjksProDE9Opi6t3GtZXHq7FTzDVOVyRdWV/DJlvZPQJfr4/ZhZlGPP xWuZw1zFUm8xbkJxMz3TfVTAtEiTbS9cpQ3oxEMibrILB5slI5k42h9PrjvQFHnHbY/Z AXs+DRzPjr4BKE2eyICJ0QHkpNrV/RXs9QqFBPf05gm/tCep3IlhoXlQL4q6mV11Y+2W cdB1NnVNlf9LJqNxdtkoP32EIvrVXSx9n7kfCALQN9EY0d24sP3v4prYkolCw/p47HNb 1uQg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=3/Dl5PXNfiRS8xSHTBcU7A9OZpqtcxgTL6VyLjhU22w=; b=cmr3s4Tc+ZqJoRXQTkJzUSrXyE0TfQ42VMiEPeK1f2ECKcAYUSpmHtBgPT5YYc3suN dN/n0sHXkJyFd9CPLNh3400MFfH4S7UaTBviAVU4MxL6msWU8/eUnc4LpRthGiCEyuRd 9pxvKIWHEAhQxP53AfL7TR+ivXrY9rm+6BhyhA1XxThkVQ+pQtnE31uBzxU1StSl9/zy 5qvWvGJRgIfUnY8tvw4J70uJnffH+oG9W3dnn4vtJ620ljixbwPZ8mJWjl/Zz29FjHrx SGonym6V2CIhYmZ4bhMOp90LT1DLW45sHSuLmT0wa3KaJ7t6Vu4ANzokrajHkoSs8YQg c3NQ== X-Gm-Message-State: AOAM530cl/DdfMKzALe9xpZUSQVffqbc3cJ9Q4gbB3mqCfJGe7/cRpYR 4NXl4DbsHA4honVZLmdUZA6V7oifbYHBvQ== X-Google-Smtp-Source: ABdhPJwBOGAhmtHVptkZb8wTHJ3jc01J3kBHJsbR4JLTLl5GT/BvjAlGEkjY7CecyPPWmWbT93E7bw== X-Received: by 2002:a1c:e043:: with SMTP id x64mr1493929wmg.48.1611601749012; Mon, 25 Jan 2021 11:09:09 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Julien Grall Subject: [PATCH V5 15/22] xen/arm: Call vcpu_ioreq_handle_completion() in check_for_vcpu_work() Date: Mon, 25 Jan 2021 21:08:22 +0200 Message-Id: <1611601709-28361-16-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko This patch adds remaining bits needed for the IOREQ support on Arm. Besides just calling vcpu_ioreq_handle_completion() we need to handle it's return value to make sure that all the vCPU works are done before we return to the guest (the vcpu_ioreq_handle_completion() may return false if there is vCPU work to do or IOREQ state is invalid). For that reason we use an unbounded loop in leave_hypervisor_to_guest(). The worse that can happen here if the vCPU will never run again (the I/O will never complete). But, in Xen case, if the I/O never completes then it most likely means that something went horribly wrong with the Device Emulator. And it is most likely not safe to continue. So letting the vCPU to spin forever if the I/O never completes is a safer action than letting it continue and leaving the guest in unclear state and is the best what we can do for now. Please note, using this loop we will not spin forever on a pCPU, preventing any other vCPUs from being scheduled. At every loop we will call check_for_pcpu_work() that will process pending softirqs. In case of failure, the guest will crash and the vCPU will be unscheduled. In normal case, if the rescheduling is necessary the vCPU will be rescheduled to give place to someone else. Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Reviewed-by: Stefano Stabellini [On Arm only] Tested-by: Wei Chen Acked-by: Julien Grall --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V1 -> V2: - new patch, changes were derived from (+ new explanation): arm/ioreq: Introduce arch specific bits for IOREQ/DM features Changes V2 -> V3: - update patch description Changes V3 -> V4: - update patch description and comment in code Changes V4 -> V5: - add Stefano's R-b - update patch subject/description and comment in code, was "xen/arm: Stick around in leave_hypervisor_to_guest until I/O has completed" - change loop logic a bit - squash with changes to check_for_vcpu_work() from patch #14 --- --- xen/arch/arm/traps.c | 26 +++++++++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c index b0cd8f9..2039ff5 100644 --- a/xen/arch/arm/traps.c +++ b/xen/arch/arm/traps.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -2261,12 +2262,23 @@ static void check_for_pcpu_work(void) * Process pending work for the vCPU. Any call should be fast or * implement preemption. */ -static void check_for_vcpu_work(void) +static bool check_for_vcpu_work(void) { struct vcpu *v = current; +#ifdef CONFIG_IOREQ_SERVER + bool handled; + + local_irq_enable(); + handled = vcpu_ioreq_handle_completion(v); + local_irq_disable(); + + if ( !handled ) + return true; +#endif + if ( likely(!v->arch.need_flush_to_ram) ) - return; + return false; /* * Give a chance for the pCPU to process work before handling the vCPU @@ -2277,6 +2289,8 @@ static void check_for_vcpu_work(void) local_irq_enable(); p2m_flush_vm(v); local_irq_disable(); + + return false; } /* @@ -2289,7 +2303,13 @@ void leave_hypervisor_to_guest(void) { local_irq_disable(); - check_for_vcpu_work(); + /* + * check_for_vcpu_work() may return true if there are more work to before + * the vCPU can safely resume. This gives us an opportunity to deschedule + * the vCPU if needed. + */ + while ( check_for_vcpu_work() ) + check_for_pcpu_work(); check_for_pcpu_work(); vgic_sync_to_lrs(); From patchwork Mon Jan 25 19:08:23 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044085 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3E491C433DB for ; Mon, 25 Jan 2021 19:20:09 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id CA73B20E65 for ; Mon, 25 Jan 2021 19:20:08 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org CA73B20E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74351.133658 (Exim 4.92) (envelope-from ) id 1l47P6-00028m-KH; Mon, 25 Jan 2021 19:20:00 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74351.133658; Mon, 25 Jan 2021 19:20:00 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P6-00028G-7h; Mon, 25 Jan 2021 19:20:00 +0000 Received: by outflank-mailman (input) for mailman id 74351; Mon, 25 Jan 2021 19:19:58 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Fk-0008N7-II for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:20 +0000 Received: from mail-wm1-x332.google.com (unknown [2a00:1450:4864:20::332]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id aec5e633-8ffd-447f-9585-487aa7b11674; Mon, 25 Jan 2021 19:09:11 +0000 (UTC) Received: by mail-wm1-x332.google.com with SMTP id u14so375071wml.4 for ; Mon, 25 Jan 2021 11:09:11 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.09 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:09 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: aec5e633-8ffd-447f-9585-487aa7b11674 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=7SvefVtoov3/IYd8ENf6jEXsSOFoAD3E0ZEff2wknCs=; b=mBUydP58s6O11knRIh9GaClTISikPopHnraU1Y/+lQd452pARl6uzQMG/bGTdfJJue Qj7prZyu3+v8lFreXu1p4lx0f2VtdGmZFIC5l3TwQ1pCZL0QiGGXh9LXnaBjSdlw/Eag PbJNHX9sUVDxNjG85x5XNJqazsXqPNxddrj/1UCWYsUMhl3yM7fcL7Act/xadhrBkok6 1wukJqf6/USUwG7BB18me+gN0DKdjbS5lsezE7i2nATasFHtOAhYANnKK1fynjIZygMw 9RFviYmwNsZ4+YqnJvbXaUs1czGQ9H66dNBkfJ+Qlc8M0fm4f4J27LWAbg+3MT6b/OL9 TJug== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=7SvefVtoov3/IYd8ENf6jEXsSOFoAD3E0ZEff2wknCs=; b=mAfWx7oZQNQ78VLIfSJXI1JBVvSCmTq6HdCFYuj5puT08kxFLPrt7fHU7FrgSBq2dI MjFP1VVR1WTcOmSpwZt4m1pasrWUpjH9RFpc3bENov5hF2EDIHtCisGOhQt8Ad1FY/j6 RqpmGF04XocrEvdAlq5PB9L5K5a5jJFndV6NqaVTduwNo4w903z3FttKNfZFb8fjp6jB 8GBAfQj4aiz+JIVqOu4oWwo0CsFbiFrEzE4LxpBnRjuv3MXXcZhuRHkud6mK/3xViA0D E20/UtUeXK7VDpkyj3WKxYGJWchC214nmy8D+acRJoPK4uyWUdNzZHJ6qNEfNsbBSlL9 DArw== X-Gm-Message-State: AOAM5300haZ4Yg2parnN9A4hc2Q010YBQ23BmQuMKOX5Ij3udJuoTnZs btLvkZJIlpcoLAaVRu+HGhUxlazH1FDQAQ== X-Google-Smtp-Source: ABdhPJwEJ3gZIfa/8OZTcxKbH/KmH4yNDfUWBLCYrpeGkULxZKK7TGhR0l0RxJfPpDDWj4JPIxjnIg== X-Received: by 2002:a1c:96d7:: with SMTP id y206mr1459838wmd.9.1611601750081; Mon, 25 Jan 2021 11:09:10 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Andrew Cooper , George Dunlap , Ian Jackson , Jan Beulich , Wei Liu , =?utf-8?q?Roger_Pau_?= =?utf-8?q?Monn=C3=A9?= , Julien Grall Subject: [PATCH V5 16/22] xen/mm: Handle properly reference in set_foreign_p2m_entry() on Arm Date: Mon, 25 Jan 2021 21:08:23 +0200 Message-Id: <1611601709-28361-17-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> MIME-Version: 1.0 From: Oleksandr Tyshchenko This patch implements reference counting of foreign entries in in set_foreign_p2m_entry() on Arm. This is a mandatory action if we want to run emulator (IOREQ server) in other than dom0 domain, as we can't trust it to do the right thing if it is not running in dom0. So we need to grab a reference on the page to avoid it disappearing. It is valid to always pass "p2m_map_foreign_rw" type to guest_physmap_add_entry() since the current and foreign domains would be always different. A case when they are equal would be rejected by rcu_lock_remote_domain_by_id(). Besides the similar comment in the code put a respective ASSERT() to catch incorrect usage in future. It was tested with IOREQ feature to confirm that all the pages given to this function belong to a domain, so we can use the same approach as for XENMAPSPACE_gmfn_foreign handling in xenmem_add_to_physmap_one(). This involves adding an extra parameter for the foreign domain to set_foreign_p2m_entry() and a helper to indicate whether the arch supports the reference counting of foreign entries and the restriction for the hardware domain in the common code can be skipped for it. Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Acked-by: Stefano Stabellini Reviewed-by: Julien Grall Reviewed-by: Jan Beulich [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch, was split from: "[RFC PATCH V1 04/12] xen/arm: Introduce arch specific bits for IOREQ/DM features" - rewrite a logic to handle properly reference in set_foreign_p2m_entry() instead of treating foreign entries as p2m_ram_rw Changes V1 -> V2: - rebase according to the recent changes to acquire_resource() - update patch description - introduce arch_refcounts_p2m() - add an explanation why p2m_map_foreign_rw is valid - move set_foreign_p2m_entry() to p2m-common.h - add const to new parameter Changes V2 -> V3: - update patch description - rename arch_refcounts_p2m() to arch_acquire_resource_check() - move comment to x86’s arch_acquire_resource_check() - return rc in Arm's set_foreign_p2m_entry() - put a respective ASSERT() into Arm's set_foreign_p2m_entry() Changes V3 -> V4: - update arch_acquire_resource_check() implementation on x86 and common code which uses it, pass struct domain to the function - put ASSERT() to x86/Arm set_foreign_p2m_entry() - use arch_acquire_resource_check() in p2m_add_foreign() instead of open-coding it Changes V4 -> V5: - update x86's arch_acquire_resource_check() - use single return statement - update comment - add Jan's and Julien's R-b, Stefano's A-b --- --- xen/arch/arm/p2m.c | 26 ++++++++++++++++++++++++++ xen/arch/x86/mm/p2m.c | 9 ++++++--- xen/common/memory.c | 9 ++------- xen/include/asm-arm/p2m.h | 19 +++++++++---------- xen/include/asm-x86/p2m.h | 14 +++++++++++--- xen/include/xen/p2m-common.h | 4 ++++ 6 files changed, 58 insertions(+), 23 deletions(-) diff --git a/xen/arch/arm/p2m.c b/xen/arch/arm/p2m.c index 4eeb867..d41c4fa 100644 --- a/xen/arch/arm/p2m.c +++ b/xen/arch/arm/p2m.c @@ -1380,6 +1380,32 @@ int guest_physmap_remove_page(struct domain *d, gfn_t gfn, mfn_t mfn, return p2m_remove_mapping(d, gfn, (1 << page_order), mfn); } +int set_foreign_p2m_entry(struct domain *d, const struct domain *fd, + unsigned long gfn, mfn_t mfn) +{ + struct page_info *page = mfn_to_page(mfn); + int rc; + + ASSERT(arch_acquire_resource_check(d)); + + if ( !get_page(page, fd) ) + return -EINVAL; + + /* + * It is valid to always use p2m_map_foreign_rw here as if this gets + * called then d != fd. A case when d == fd would be rejected by + * rcu_lock_remote_domain_by_id() earlier. Put a respective ASSERT() + * to catch incorrect usage in future. + */ + ASSERT(d != fd); + + rc = guest_physmap_add_entry(d, _gfn(gfn), mfn, 0, p2m_map_foreign_rw); + if ( rc ) + put_page(page); + + return rc; +} + static struct page_info *p2m_allocate_root(void) { struct page_info *page; diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c index c1dd45b..2091aed 100644 --- a/xen/arch/x86/mm/p2m.c +++ b/xen/arch/x86/mm/p2m.c @@ -1323,8 +1323,11 @@ static int set_typed_p2m_entry(struct domain *d, unsigned long gfn_l, } /* Set foreign mfn in the given guest's p2m table. */ -int set_foreign_p2m_entry(struct domain *d, unsigned long gfn, mfn_t mfn) +int set_foreign_p2m_entry(struct domain *d, const struct domain *fd, + unsigned long gfn, mfn_t mfn) { + ASSERT(arch_acquire_resource_check(d)); + return set_typed_p2m_entry(d, gfn, mfn, PAGE_ORDER_4K, p2m_map_foreign, p2m_get_hostp2m(d)->default_access); } @@ -2587,7 +2590,7 @@ static int p2m_add_foreign(struct domain *tdom, unsigned long fgfn, * hvm fixme: until support is added to p2m teardown code to cleanup any * foreign entries, limit this to hardware domain only. */ - if ( !is_hardware_domain(tdom) ) + if ( !arch_acquire_resource_check(tdom) ) return -EPERM; if ( foreigndom == DOMID_XEN ) @@ -2643,7 +2646,7 @@ static int p2m_add_foreign(struct domain *tdom, unsigned long fgfn, * will update the m2p table which will result in mfn -> gpfn of dom0 * and not fgfn of domU. */ - rc = set_foreign_p2m_entry(tdom, gpfn, mfn); + rc = set_foreign_p2m_entry(tdom, fdom, gpfn, mfn); if ( rc ) gdprintk(XENLOG_WARNING, "set_foreign_p2m_entry failed. " "gpfn:%lx mfn:%lx fgfn:%lx td:%d fd:%d\n", diff --git a/xen/common/memory.c b/xen/common/memory.c index 8eb05b1..33296e6 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -1139,12 +1139,7 @@ static int acquire_resource( xen_pfn_t mfn_list[32]; int rc; - /* - * FIXME: Until foreign pages inserted into the P2M are properly - * reference counted, it is unsafe to allow mapping of - * resource pages unless the caller is the hardware domain. - */ - if ( paging_mode_translate(currd) && !is_hardware_domain(currd) ) + if ( !arch_acquire_resource_check(currd) ) return -EACCES; if ( copy_from_guest(&xmar, arg, 1) ) @@ -1212,7 +1207,7 @@ static int acquire_resource( for ( i = 0; !rc && i < xmar.nr_frames; i++ ) { - rc = set_foreign_p2m_entry(currd, gfn_list[i], + rc = set_foreign_p2m_entry(currd, d, gfn_list[i], _mfn(mfn_list[i])); /* rc should be -EIO for any iteration other than the first */ if ( rc && i ) diff --git a/xen/include/asm-arm/p2m.h b/xen/include/asm-arm/p2m.h index 28ca9a8..4f8b3b0 100644 --- a/xen/include/asm-arm/p2m.h +++ b/xen/include/asm-arm/p2m.h @@ -161,6 +161,15 @@ typedef enum { #endif #include +static inline bool arch_acquire_resource_check(struct domain *d) +{ + /* + * The reference counting of foreign entries in set_foreign_p2m_entry() + * is supported on Arm. + */ + return true; +} + static inline void p2m_altp2m_check(struct vcpu *v, uint16_t idx) { @@ -392,16 +401,6 @@ static inline gfn_t gfn_next_boundary(gfn_t gfn, unsigned int order) return gfn_add(gfn, 1UL << order); } -static inline int set_foreign_p2m_entry(struct domain *d, unsigned long gfn, - mfn_t mfn) -{ - /* - * NOTE: If this is implemented then proper reference counting of - * foreign entries will need to be implemented. - */ - return -EOPNOTSUPP; -} - /* * A vCPU has cache enabled only when the MMU is enabled and data cache * is enabled. diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h index 5d7836d..7d63f57 100644 --- a/xen/include/asm-x86/p2m.h +++ b/xen/include/asm-x86/p2m.h @@ -382,6 +382,17 @@ struct p2m_domain { #endif #include +static inline bool arch_acquire_resource_check(struct domain *d) +{ + /* + * FIXME: Until foreign pages inserted into the P2M are properly + * reference counted, it is unsafe to allow mapping of + * resource pages unless the caller is the hardware domain + * (see set_foreign_p2m_entry()). + */ + return !paging_mode_translate(d) || is_hardware_domain(d); +} + /* * Updates vCPU's n2pm to match its np2m_base in VMCx12 and returns that np2m. */ @@ -647,9 +658,6 @@ int p2m_finish_type_change(struct domain *d, int p2m_is_logdirty_range(struct p2m_domain *, unsigned long start, unsigned long end); -/* Set foreign entry in the p2m table (for priv-mapping) */ -int set_foreign_p2m_entry(struct domain *d, unsigned long gfn, mfn_t mfn); - /* Set mmio addresses in the p2m table (for pass-through) */ int set_mmio_p2m_entry(struct domain *d, gfn_t gfn, mfn_t mfn, unsigned int order); diff --git a/xen/include/xen/p2m-common.h b/xen/include/xen/p2m-common.h index 3753bc0..a322e73 100644 --- a/xen/include/xen/p2m-common.h +++ b/xen/include/xen/p2m-common.h @@ -3,6 +3,10 @@ #include +/* Set foreign entry in the p2m table */ +int set_foreign_p2m_entry(struct domain *d, const struct domain *fd, + unsigned long gfn, mfn_t mfn); + /* Remove a page from a domain's p2m table */ int __must_check guest_physmap_remove_page(struct domain *d, gfn_t gfn, mfn_t mfn, From patchwork Mon Jan 25 19:08:24 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044083 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 2B2D6C433E0 for ; Mon, 25 Jan 2021 19:20:07 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id D2BB722525 for ; Mon, 25 Jan 2021 19:20:06 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org D2BB722525 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74349.133645 (Exim 4.92) (envelope-from ) id 1l47P5-00026w-OV; Mon, 25 Jan 2021 19:19:59 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74349.133645; Mon, 25 Jan 2021 19:19:59 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P5-00026n-JY; Mon, 25 Jan 2021 19:19:59 +0000 Received: by outflank-mailman (input) for mailman id 74349; Mon, 25 Jan 2021 19:19:57 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Fp-0008N7-IV for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:25 +0000 Received: from mail-wm1-x330.google.com (unknown [2a00:1450:4864:20::330]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 82ac12d1-e0c8-431d-9c79-6bb92b3af79a; Mon, 25 Jan 2021 19:09:12 +0000 (UTC) Received: by mail-wm1-x330.google.com with SMTP id i9so561369wmq.1 for ; Mon, 25 Jan 2021 11:09:11 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.10 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:10 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 82ac12d1-e0c8-431d-9c79-6bb92b3af79a DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=40mdufwi456sOAvEGGJAfww9prSnsvb2RITUqz5aqvQ=; b=LTDketEEaOUlxBeDbQvhfdqx2EGE5X723UzuEsQPvowoSn7yWYLonQ/QTiAPmv7t6i 7ntKYatrlFvtmeU5HoGHR4pI0r6kuOboENd3Qu9UiiMesbWprr4KHEDTWePS06Jj4+a5 Jb1ysiNcbJgpopKarFa8Zzv1yzalCSTlbRg8kJ2VvgcAfS+QsWxMchS4METbiYxHzKFM jYXG3olMzqBZU2ww4bKNzHmbBhsnti41Ouiz21ndIm73QTaC5hM6/3CTiCxfTbxD4Yxa pmti0h78Gu7/TiFYiG0Pdutrb0neklTHJAqNfTwhuIH2WsIUBYnxNk5PSLe5GBV91sNA rxcQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=40mdufwi456sOAvEGGJAfww9prSnsvb2RITUqz5aqvQ=; b=BgHQV2BYKJegjWEacc9QGh8CGm4IEadPaFul8QKDCKTeDLJDT9Th4Pa+Np9LZJAkjh 9Cc9UW0p/Dovq+oUontxictcI207t3qMKRuhrK4QEoKE4G9EWihSeG1IGW8eg9+durAV L5jzlePJmmb9Kbe9/ZC46WCjkQ6FMedzlqGDEtWeVVxdZrYFlrEIrRSFkmHAQGGewZ4R rgQnj4Yyl4YcXNIKtaSSJaMFQAjUv+1zTHzigp1noqrvpFGFJwwi5DM1AK+eSUwalOyB AfMsmQGEkePKFKt9oGLrm482CjWrxxdzFC//45iI3C8Jbor4GkAwaup07Fcex6KvMQZb fVaQ== X-Gm-Message-State: AOAM533M1QeaPQBvkzya5wfRtxVXtrR2dyvYwGXZdNlzZ+KtpAOgsedf 94v1dTMP6V55EQgByXsXAb7GiNkRlJz/eg== X-Google-Smtp-Source: ABdhPJyC20Bu0VXZpG/L4U6pzWaBuR/eBS9UfA3ZYfw6wRr8mjrSTsSfebjSNqqjX6IMIFbXzyueDg== X-Received: by 2002:a1c:b7d6:: with SMTP id h205mr1468426wmf.182.1611601751070; Mon, 25 Jan 2021 11:09:11 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Paul Durrant , Julien Grall Subject: [PATCH V5 17/22] xen/ioreq: Introduce domain_has_ioreq_server() Date: Mon, 25 Jan 2021 21:08:24 +0200 Message-Id: <1611601709-28361-18-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko This patch introduces a helper the main purpose of which is to check if a domain is using IOREQ server(s). On Arm the current benefit is to avoid calling vcpu_ioreq_handle_completion() (which implies iterating over all possible IOREQ servers anyway) on every return in leave_hypervisor_to_guest() if there is no active servers for the particular domain. Also this helper will be used by one of the subsequent patches on Arm. Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Reviewed-by: Stefano Stabellini Reviewed-by: Paul Durrant [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - new patch Changes V1 -> V2: - update patch description - guard helper with CONFIG_IOREQ_SERVER - remove "hvm" prefix - modify helper to just return d->arch.hvm.ioreq_server.nr_servers - put suitable ASSERT()s - use ASSERT(d->ioreq_server.server[id] ? !s : !!s) in set_ioreq_server() - remove d->ioreq_server.nr_servers = 0 from hvm_ioreq_init() Changes V2 -> V3: - update patch description - remove ASSERT()s from the helper, add a comment - use #ifdef CONFIG_IOREQ_SERVER inside function body - use new ASSERT() construction in set_ioreq_server() Changes V3 -> V4: - update patch description - drop per-domain variable "nr_servers" - reimplement a helper to count the non-NULL entries - make the helper out-of-line Changes V4 -> V5: - add Stefano's and Paul's R-b --- --- xen/arch/arm/traps.c | 15 +++++++++------ xen/common/ioreq.c | 16 ++++++++++++++++ xen/include/xen/ioreq.h | 2 ++ 3 files changed, 27 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c index 2039ff5..4cdd343 100644 --- a/xen/arch/arm/traps.c +++ b/xen/arch/arm/traps.c @@ -2267,14 +2267,17 @@ static bool check_for_vcpu_work(void) struct vcpu *v = current; #ifdef CONFIG_IOREQ_SERVER - bool handled; + if ( domain_has_ioreq_server(v->domain) ) + { + bool handled; - local_irq_enable(); - handled = vcpu_ioreq_handle_completion(v); - local_irq_disable(); + local_irq_enable(); + handled = vcpu_ioreq_handle_completion(v); + local_irq_disable(); - if ( !handled ) - return true; + if ( !handled ) + return true; + } #endif if ( likely(!v->arch.need_flush_to_ram) ) diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 07572a5..5b0f03e 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -80,6 +80,22 @@ static ioreq_t *get_ioreq(struct ioreq_server *s, struct vcpu *v) return &p->vcpu_ioreq[v->vcpu_id]; } +/* + * This should only be used when d == current->domain or when they're + * distinct and d is paused. Otherwise the result is stale before + * the caller can inspect it. + */ +bool domain_has_ioreq_server(const struct domain *d) +{ + const struct ioreq_server *s; + unsigned int id; + + FOR_EACH_IOREQ_SERVER(d, id, s) + return true; + + return false; +} + static struct ioreq_vcpu *get_pending_vcpu(const struct vcpu *v, struct ioreq_server **srvp) { diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 0b433e2..89ee171 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -83,6 +83,8 @@ static inline bool ioreq_needs_completion(const ioreq_t *ioreq) #define HANDLE_BUFIOREQ(s) \ ((s)->bufioreq_handling != HVM_IOREQSRV_BUFIOREQ_OFF) +bool domain_has_ioreq_server(const struct domain *d); + bool vcpu_ioreq_pending(struct vcpu *v); bool vcpu_ioreq_handle_completion(struct vcpu *v); bool is_ioreq_server_page(struct domain *d, const struct page_info *page); From patchwork Mon Jan 25 19:08:25 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044099 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 23FE0C433E0 for ; Mon, 25 Jan 2021 19:21:20 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id BA84721744 for ; Mon, 25 Jan 2021 19:21:19 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org BA84721744 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74383.133770 (Exim 4.92) (envelope-from ) id 1l47QE-0003k8-HZ; Mon, 25 Jan 2021 19:21:10 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74383.133770; Mon, 25 Jan 2021 19:21:10 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47QE-0003jF-1R; Mon, 25 Jan 2021 19:21:10 +0000 Received: by outflank-mailman (input) for mailman id 74383; Mon, 25 Jan 2021 19:21:07 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Fz-0008N7-Ip for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:35 +0000 Received: from mail-wm1-x32e.google.com (unknown [2a00:1450:4864:20::32e]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 4677ca31-8225-4649-b648-8bcdf5dd0cf6; Mon, 25 Jan 2021 19:09:14 +0000 (UTC) Received: by mail-wm1-x32e.google.com with SMTP id 190so405220wmz.0 for ; Mon, 25 Jan 2021 11:09:13 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.11 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:11 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 4677ca31-8225-4649-b648-8bcdf5dd0cf6 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=SCUpLEiOlSX9ul+kRijfipf8pIcb5kXp2unnhpR7jAc=; b=prv5iMM/uS0Q196uQKNNJjf/gKceib37WdsQCJHp1q2vDrr0xOoCYw6wMVg088uHpu KrS41/3QQ4rVgZ6uYcAO9voy3nnP+eCgfd29AU5jtFUxC+iuvBtXtozhXtTGUjxD4By5 U54Nb1T1ojEUPL1olsiLxpOh4mqF2GBTHuF1qcU+irE/cGsSLEZwQV6TPfdRRDfoXTma 6VN1fMDRIlEyBh739WOE4E9nez5XRqOzFt0+Cyz2zWMOs+EbYaslvtwDzrDyK0GtiUuK wSuBVGshVAJ2Je/SV2tVz52eHKAcj3VzKuePUYUorPbrBg4/J8dvk43YGDFpvTRaGP2x QCtw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=SCUpLEiOlSX9ul+kRijfipf8pIcb5kXp2unnhpR7jAc=; b=bS0JgaWur6U9oCPUaOe3MslSe6EPbyuqApgnQgDDRs5UtiVxu8U+ch4x6wVDb4ndyj apMMwLr3LjVSgjVPIpdg07L5JBAOS2j67Uh4nfSpq/FBtku2yB97IDTqW9+4hzSqpdnA RITelx1nfyDHoehxUjN3vs57Vwj3pzMVWBKaUT6hd1qPLZA3PWvDchPgIA0q7uy9CkuR t76Sc6OkMoOeYg59H9Iwxjs50TOrZSF/pj4Ee5PNdl78OrHPMkOR8aLOfLzbC+u6yBKf X0tVe7MgRvHNYPA/bA28GOX41ii7x2N1VWshKXOdTdDaG/PHBF5caGUCeKAWszEF/jU5 1Pfg== X-Gm-Message-State: AOAM5338X7zYmt7l44vDeefUNElh9x5MmAQP6w2l9rHERvjzqupft7Jk YCaLOXa5YpMQ7u3/L43HzF/l/jF/8pMLqg== X-Google-Smtp-Source: ABdhPJzCrjr+H6hjoOH0LCGPFQHL0w7RQK2ptp43b+vUuwWK1ofTkk2w2FoTTyY3udPjkYQy0bRUeA== X-Received: by 2002:a1c:398b:: with SMTP id g133mr1416143wma.35.1611601752109; Mon, 25 Jan 2021 11:09:12 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Ian Jackson , Wei Liu , Andrew Cooper , George Dunlap , Jan Beulich , Julien Grall , Stefano Stabellini , Volodymyr Babchuk , Julien Grall Subject: [PATCH V5 18/22] xen/dm: Introduce xendevicemodel_set_irq_level DM op Date: Mon, 25 Jan 2021 21:08:25 +0200 Message-Id: <1611601709-28361-19-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko This patch adds ability to the device emulator to notify otherend (some entity running in the guest) using a SPI and implements Arm specific bits for it. Proposed interface allows emulator to set the logical level of a one of a domain's IRQ lines. We can't reuse the existing DM op (xen_dm_op_set_isa_irq_level) to inject an interrupt as the "isa_irq" field is only 8-bit and able to cover IRQ 0 - 255, whereas we need a wider range (0 - 1020). Please note, for egde-triggered interrupt (which is used for the virtio-mmio emulation) we only trigger the interrupt on Arm if the level is asserted (rising edge) and do nothing if the level is deasserted (falling edge), so the call could be named "trigger_irq" (without the level parameter). But, in order to model the line closely (to be able to support level-triggered interrupt) we need to know whether the line is low or high, so the proposed interface has been chosen. However, it is worth mentioning that in case of the level-triggered interrupt, we should keep injecting the interrupt to the guest until the line is deasserted (this is not covered by current patch). Signed-off-by: Julien Grall Signed-off-by: Oleksandr Tyshchenko Acked-by: Stefano Stabellini [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - check incoming parameters in arch_dm_op() - add explicit padding to struct xen_dm_op_set_irq_level Changes V1 -> V2: - update the author of a patch - update patch description - check that padding is always 0 - mention that interface is Arm only and only SPIs are supported for now - allow to set the logical level of a line for non-allocated interrupts only - add xen_dm_op_set_irq_level_t Changes V2 -> V3: - no changes Changes V3 -> V4: - update patch description - update patch according to the IOREQ related dm-op handling changes Changes V4 -> V5: - rebase - add Stefano-s A-b --- --- tools/include/xendevicemodel.h | 4 +++ tools/libs/devicemodel/core.c | 18 ++++++++++ tools/libs/devicemodel/libxendevicemodel.map | 1 + xen/arch/arm/dm.c | 54 +++++++++++++++++++++++++++- xen/include/public/hvm/dm_op.h | 16 +++++++++ 5 files changed, 92 insertions(+), 1 deletion(-) diff --git a/tools/include/xendevicemodel.h b/tools/include/xendevicemodel.h index e877f5c..c06b3c8 100644 --- a/tools/include/xendevicemodel.h +++ b/tools/include/xendevicemodel.h @@ -209,6 +209,10 @@ int xendevicemodel_set_isa_irq_level( xendevicemodel_handle *dmod, domid_t domid, uint8_t irq, unsigned int level); +int xendevicemodel_set_irq_level( + xendevicemodel_handle *dmod, domid_t domid, unsigned int irq, + unsigned int level); + /** * This function maps a PCI INTx line to a an IRQ line. * diff --git a/tools/libs/devicemodel/core.c b/tools/libs/devicemodel/core.c index 4d40639..30bd79f 100644 --- a/tools/libs/devicemodel/core.c +++ b/tools/libs/devicemodel/core.c @@ -430,6 +430,24 @@ int xendevicemodel_set_isa_irq_level( return xendevicemodel_op(dmod, domid, 1, &op, sizeof(op)); } +int xendevicemodel_set_irq_level( + xendevicemodel_handle *dmod, domid_t domid, uint32_t irq, + unsigned int level) +{ + struct xen_dm_op op; + struct xen_dm_op_set_irq_level *data; + + memset(&op, 0, sizeof(op)); + + op.op = XEN_DMOP_set_irq_level; + data = &op.u.set_irq_level; + + data->irq = irq; + data->level = level; + + return xendevicemodel_op(dmod, domid, 1, &op, sizeof(op)); +} + int xendevicemodel_set_pci_link_route( xendevicemodel_handle *dmod, domid_t domid, uint8_t link, uint8_t irq) { diff --git a/tools/libs/devicemodel/libxendevicemodel.map b/tools/libs/devicemodel/libxendevicemodel.map index 561c62d..a0c3012 100644 --- a/tools/libs/devicemodel/libxendevicemodel.map +++ b/tools/libs/devicemodel/libxendevicemodel.map @@ -32,6 +32,7 @@ VERS_1.2 { global: xendevicemodel_relocate_memory; xendevicemodel_pin_memory_cacheattr; + xendevicemodel_set_irq_level; } VERS_1.1; VERS_1.3 { diff --git a/xen/arch/arm/dm.c b/xen/arch/arm/dm.c index f254ed7..7854133 100644 --- a/xen/arch/arm/dm.c +++ b/xen/arch/arm/dm.c @@ -20,6 +20,8 @@ #include #include +#include + int dm_op(const struct dmop_args *op_args) { struct domain *d; @@ -35,6 +37,7 @@ int dm_op(const struct dmop_args *op_args) [XEN_DMOP_unmap_io_range_from_ioreq_server] = sizeof(struct xen_dm_op_ioreq_server_range), [XEN_DMOP_set_ioreq_server_state] = sizeof(struct xen_dm_op_set_ioreq_server_state), [XEN_DMOP_destroy_ioreq_server] = sizeof(struct xen_dm_op_destroy_ioreq_server), + [XEN_DMOP_set_irq_level] = sizeof(struct xen_dm_op_set_irq_level), }; rc = rcu_lock_remote_domain_by_id(op_args->domid, &d); @@ -73,7 +76,56 @@ int dm_op(const struct dmop_args *op_args) if ( op.pad ) goto out; - rc = ioreq_server_dm_op(&op, d, &const_op); + switch ( op.op ) + { + case XEN_DMOP_set_irq_level: + { + const struct xen_dm_op_set_irq_level *data = + &op.u.set_irq_level; + unsigned int i; + + /* Only SPIs are supported */ + if ( (data->irq < NR_LOCAL_IRQS) || (data->irq >= vgic_num_irqs(d)) ) + { + rc = -EINVAL; + break; + } + + if ( data->level != 0 && data->level != 1 ) + { + rc = -EINVAL; + break; + } + + /* Check that padding is always 0 */ + for ( i = 0; i < sizeof(data->pad); i++ ) + { + if ( data->pad[i] ) + { + rc = -EINVAL; + break; + } + } + + /* + * Allow to set the logical level of a line for non-allocated + * interrupts only. + */ + if ( test_bit(data->irq, d->arch.vgic.allocated_irqs) ) + { + rc = -EINVAL; + break; + } + + vgic_inject_irq(d, NULL, data->irq, data->level); + rc = 0; + break; + } + + default: + rc = ioreq_server_dm_op(&op, d, &const_op); + break; + } if ( (!rc || rc == -ERESTART) && !const_op && copy_to_guest_offset(op_args->buf[0].h, offset, diff --git a/xen/include/public/hvm/dm_op.h b/xen/include/public/hvm/dm_op.h index 66cae1a..1f70d58 100644 --- a/xen/include/public/hvm/dm_op.h +++ b/xen/include/public/hvm/dm_op.h @@ -434,6 +434,21 @@ struct xen_dm_op_pin_memory_cacheattr { }; typedef struct xen_dm_op_pin_memory_cacheattr xen_dm_op_pin_memory_cacheattr_t; +/* + * XEN_DMOP_set_irq_level: Set the logical level of a one of a domain's + * IRQ lines (currently Arm only). + * Only SPIs are supported. + */ +#define XEN_DMOP_set_irq_level 19 + +struct xen_dm_op_set_irq_level { + uint32_t irq; + /* IN - Level: 0 -> deasserted, 1 -> asserted */ + uint8_t level; + uint8_t pad[3]; +}; +typedef struct xen_dm_op_set_irq_level xen_dm_op_set_irq_level_t; + struct xen_dm_op { uint32_t op; uint32_t pad; @@ -447,6 +462,7 @@ struct xen_dm_op { xen_dm_op_track_dirty_vram_t track_dirty_vram; xen_dm_op_set_pci_intx_level_t set_pci_intx_level; xen_dm_op_set_isa_irq_level_t set_isa_irq_level; + xen_dm_op_set_irq_level_t set_irq_level; xen_dm_op_set_pci_link_route_t set_pci_link_route; xen_dm_op_modified_memory_t modified_memory; xen_dm_op_set_mem_type_t set_mem_type; From patchwork Mon Jan 25 19:08:26 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044087 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 7AE84C43381 for ; Mon, 25 Jan 2021 19:20:09 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 218B821D94 for ; Mon, 25 Jan 2021 19:20:08 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 218B821D94 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74352.133671 (Exim 4.92) (envelope-from ) id 1l47P7-0002AO-BY; Mon, 25 Jan 2021 19:20:01 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74352.133671; Mon, 25 Jan 2021 19:20:01 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47P6-00029a-Ny; Mon, 25 Jan 2021 19:20:00 +0000 Received: by outflank-mailman (input) for mailman id 74352; Mon, 25 Jan 2021 19:19:58 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47Fu-0008N7-IY for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:30 +0000 Received: from mail-wm1-x32f.google.com (unknown [2a00:1450:4864:20::32f]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 4dfaaa4d-5dbd-425e-9d2c-1919a684d858; Mon, 25 Jan 2021 19:09:14 +0000 (UTC) Received: by mail-wm1-x32f.google.com with SMTP id 190so405291wmz.0 for ; Mon, 25 Jan 2021 11:09:14 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.12 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:12 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 4dfaaa4d-5dbd-425e-9d2c-1919a684d858 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=pxgov1xTpd8KSnySmM4D4uZTYyiz6I2VdpWg/8wV51k=; b=W03jV8YEA/hR0F4fHiOoCDKFawl/qg34vNIqq+NGhzipFTuuFdx7wcv+lRy7ksGeMu Nik6eCRjOZb22kpissSQ90IGVhrd/qb7KFFdjq+QrHdq1sogZerWbxPhP1t1bJYAF+/k iGR6DdduTW4uCzmeJgVzI3OTQkKh/K3A1Abz2oG34khU57tp1qNdI6D2nrZVTzpgy5tS fXh9Cwh6mq2F+RS8W1DGiQKDinVd9LPyF+VhCb/1g+RKzXt7zxIBqfpLglHFSXhPqnrr Df08V2grxXpR/MCI5O+yTIvPBCYt1EE4jdfw3q8NhEvmMAeYkIp67nYLsdywH4jOPYxb HlYw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=pxgov1xTpd8KSnySmM4D4uZTYyiz6I2VdpWg/8wV51k=; b=SFsWKbWqVf7bYygPXMNxhGU4JUdH086HXKey42jDjuVhMDDjuklArMzemLt5eMEe8t SYiESu9fsBX0lGwZdd9BJaDIKrkoplGeO9dMVLVkaXLgr6NVfISFxitAeaa6P9eVl71r Ultah475N2+Fcdpn/JvztDvBRsaDToxHg4a/WxurfEc0GQ+1mwjHgNaghOsprPhwtDQG 55XhD6dm+8QNz0mRLMZ/7IXEDBzcWy24LbQt+y+7jiQujjvzivSa3jsFZLPu6QD6PBuM zoMtCLmE8V81KR0q3y3ymj4eDkULAADmLdWOkXMAFy9ZQR6LPKvzZaoth4f0x5H+1sHL A1iA== X-Gm-Message-State: AOAM532jHuu9HYQ/9jAyraHKVseYh3oZdFHSGp8oZJ+0J5AhIj2Ks4Q3 FE2lqr3TGvGKq4WymUiBE3VPsx4jWuOs3Q== X-Google-Smtp-Source: ABdhPJwbvxnS2Vdxbms+fUo2bzss4WZ733bXU/Z9kbBbeM5k0GlYC+GRHiZ0+HOtAM9BrdGzPkzNEw== X-Received: by 2002:a1c:e912:: with SMTP id q18mr1462529wmc.162.1611601753061; Mon, 25 Jan 2021 11:09:13 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Julien Grall Subject: [PATCH V5 19/22] xen/arm: io: Abstract sign-extension Date: Mon, 25 Jan 2021 21:08:26 +0200 Message-Id: <1611601709-28361-20-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko In order to avoid code duplication (both handle_read() and handle_ioserv() contain the same code for the sign-extension) put this code to a common helper to be used for both. Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Reviewed-by: Stefano Stabellini [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V1 -> V2: - new patch Changes V2 -> V3: - no changes Changes V3 -> V4: - no changes here, but in new patch: "xen/arm: io: Harden sign extension check" Changes V4 -> V5: - add Stefano-s R-b --- --- xen/arch/arm/io.c | 18 ++---------------- xen/arch/arm/ioreq.c | 17 +---------------- xen/include/asm-arm/traps.h | 24 ++++++++++++++++++++++++ 3 files changed, 27 insertions(+), 32 deletions(-) diff --git a/xen/arch/arm/io.c b/xen/arch/arm/io.c index 7ac0303..729287e 100644 --- a/xen/arch/arm/io.c +++ b/xen/arch/arm/io.c @@ -25,6 +25,7 @@ #include #include #include +#include #include "decode.h" @@ -40,26 +41,11 @@ static enum io_state handle_read(const struct mmio_handler *handler, * setting r). */ register_t r = 0; - uint8_t size = (1 << dabt.size) * 8; if ( !handler->ops->read(v, info, &r, handler->priv) ) return IO_ABORT; - /* - * Sign extend if required. - * Note that we expect the read handler to have zeroed the bits - * outside the requested access size. - */ - if ( dabt.sign && (r & (1UL << (size - 1))) ) - { - /* - * We are relying on register_t using the same as - * an unsigned long in order to keep the 32-bit assembly - * code smaller. - */ - BUILD_BUG_ON(sizeof(register_t) != sizeof(unsigned long)); - r |= (~0UL) << size; - } + r = sign_extend(dabt, r); set_user_reg(regs, dabt.reg, r); diff --git a/xen/arch/arm/ioreq.c b/xen/arch/arm/ioreq.c index ffeeb0b..c7ee1a7 100644 --- a/xen/arch/arm/ioreq.c +++ b/xen/arch/arm/ioreq.c @@ -28,7 +28,6 @@ enum io_state handle_ioserv(struct cpu_user_regs *regs, struct vcpu *v) const union hsr hsr = { .bits = regs->hsr }; const struct hsr_dabt dabt = hsr.dabt; /* Code is similar to handle_read */ - uint8_t size = (1 << dabt.size) * 8; register_t r = v->io.req.data; /* We are done with the IO */ @@ -37,21 +36,7 @@ enum io_state handle_ioserv(struct cpu_user_regs *regs, struct vcpu *v) if ( dabt.write ) return IO_HANDLED; - /* - * Sign extend if required. - * Note that we expect the read handler to have zeroed the bits - * outside the requested access size. - */ - if ( dabt.sign && (r & (1UL << (size - 1))) ) - { - /* - * We are relying on register_t using the same as - * an unsigned long in order to keep the 32-bit assembly - * code smaller. - */ - BUILD_BUG_ON(sizeof(register_t) != sizeof(unsigned long)); - r |= (~0UL) << size; - } + r = sign_extend(dabt, r); set_user_reg(regs, dabt.reg, r); diff --git a/xen/include/asm-arm/traps.h b/xen/include/asm-arm/traps.h index c4a3d0f..c6b3cc7 100644 --- a/xen/include/asm-arm/traps.h +++ b/xen/include/asm-arm/traps.h @@ -84,6 +84,30 @@ static inline bool VABORT_GEN_BY_GUEST(const struct cpu_user_regs *regs) (unsigned long)abort_guest_exit_end == regs->pc; } +/* Check whether the sign extension is required and perform it */ +static inline register_t sign_extend(const struct hsr_dabt dabt, register_t r) +{ + uint8_t size = (1 << dabt.size) * 8; + + /* + * Sign extend if required. + * Note that we expect the read handler to have zeroed the bits + * outside the requested access size. + */ + if ( dabt.sign && (r & (1UL << (size - 1))) ) + { + /* + * We are relying on register_t using the same as + * an unsigned long in order to keep the 32-bit assembly + * code smaller. + */ + BUILD_BUG_ON(sizeof(register_t) != sizeof(unsigned long)); + r |= (~0UL) << size; + } + + return r; +} + #endif /* __ASM_ARM_TRAPS__ */ /* * Local variables: From patchwork Mon Jan 25 19:08:27 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044095 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 88D3FC433E0 for ; Mon, 25 Jan 2021 19:21:17 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 41B0A224DF for ; Mon, 25 Jan 2021 19:21:17 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 41B0A224DF Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74378.133729 (Exim 4.92) (envelope-from ) id 1l47QC-0003eb-4p; Mon, 25 Jan 2021 19:21:08 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74378.133729; Mon, 25 Jan 2021 19:21:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47QC-0003eT-1S; Mon, 25 Jan 2021 19:21:08 +0000 Received: by outflank-mailman (input) for mailman id 74378; Mon, 25 Jan 2021 19:21:06 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47G9-0008N7-JA for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:45 +0000 Received: from mail-wr1-x434.google.com (unknown [2a00:1450:4864:20::434]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 0b3dc2d8-998b-411b-a0ae-aec95ec0f12b; Mon, 25 Jan 2021 19:09:14 +0000 (UTC) Received: by mail-wr1-x434.google.com with SMTP id c12so14121099wrc.7 for ; Mon, 25 Jan 2021 11:09:14 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.13 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:13 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 0b3dc2d8-998b-411b-a0ae-aec95ec0f12b DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=Sk79uDPAYTcbZS3IWNRXkK9ttYsPF3BwQfUARr2zP5w=; b=S7bBHY2jX+8tT1AkkSi7NewjpnxqKyRltDTV6cKkz4Q1QjzsPrP9jFz8UNhlvabht5 eAcvd6DtFO00KKwee+dks2tykDGiw7rW0VGEsReEdapj96QO11rrliP48DI+BBEsDywF feYjUInrg7UR54yKc/SKixnOw+v8g0R8wetkrOfIILD6FNImBFUsOpSzFuVXCzD9/3fF L5xxR7RDedjqmBkINiJWmMnT4bCAAYAFPv+9ijqUf05/RGW4si53m6RrUosSRauizC+G s1Nts/MJc34un8xUp8INbtDBu1uIDJVSb6z1yqmivDUgwS3Cmc0xATTX3Wq0Av2iEhpN kq3w== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=Sk79uDPAYTcbZS3IWNRXkK9ttYsPF3BwQfUARr2zP5w=; b=bM2TvVfLLSpaolbEhM5gV8E7IXBlugIdCM4qIuEOgiAnJ/M+aDX854aoHREpHk7Gix 4leHm+K76qUFwYEgOoyHsVgqdWDfCS2aEGITukE4l0aXbadMo2Ghb6ubuCBBcfISCaXu yEI7OAhIcthPQTTps8Vs07BkJhju8wi3HR28EwkuOGVDbpUHewOLa/jwn2yRlVTAtvUE 4WsTir7JZfIrlv3jCoIcjZeDDjPwG7cE1N0DWs9O2UyGg+qZ/eTcKLsd677yeAmW0436 ZV4OwkUNlEQhqtq8h9ofcOBiBk5aRNihWdDqTqisDsjtCW8xrLmkd/mDFcEPeVLfq3Gg 329g== X-Gm-Message-State: AOAM530zaYe0ir2dNDdgaWIA/gQQL9BYN4YS/9NgfdSxm69z00JVz45x btt39H38J4PkipB/0Ao5eJ9JeFXq/ZMPIw== X-Google-Smtp-Source: ABdhPJwg+fou6IhlCR6vpmF0nriE79P9MuJzT14u5Hxu9FJE+RZ/uJ8CEUoUfciGBDP7VPb8+haqCw== X-Received: by 2002:a5d:6092:: with SMTP id w18mr2543200wrt.75.1611601753964; Mon, 25 Jan 2021 11:09:13 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Julien Grall Subject: [PATCH V5 20/22] xen/arm: io: Harden sign extension check Date: Mon, 25 Jan 2021 21:08:27 +0200 Message-Id: <1611601709-28361-21-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko In the ideal world we would never get an undefined behavior when propagating the sign bit since that bit can only be set for access size smaller than the register size (i.e byte/half-word for aarch32, byte/half-word/word for aarch64). In the real world we need to care for *possible* hardware bug such as advertising a sign extension for either 64-bit (or 32-bit) on Arm64 (resp. Arm32). So harden a bit more the code to prevent undefined behavior when propagating the sign bit in case of buggy hardware. Signed-off-by: Oleksandr Tyshchenko Reviewed-by: Stefano Stabellini Reviewed-by: Volodymyr Babchuk CC: Julien Grall --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes V3 -> V4: - new patch Changes V4 -> V5: - add Stefano's and Volodymyr's R-b --- --- xen/include/asm-arm/traps.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/xen/include/asm-arm/traps.h b/xen/include/asm-arm/traps.h index c6b3cc7..2ed2b85 100644 --- a/xen/include/asm-arm/traps.h +++ b/xen/include/asm-arm/traps.h @@ -94,7 +94,8 @@ static inline register_t sign_extend(const struct hsr_dabt dabt, register_t r) * Note that we expect the read handler to have zeroed the bits * outside the requested access size. */ - if ( dabt.sign && (r & (1UL << (size - 1))) ) + if ( dabt.sign && (size < sizeof(register_t) * 8) && + (r & (1UL << (size - 1))) ) { /* * We are relying on register_t using the same as From patchwork Mon Jan 25 19:08:28 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044079 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 258D5C433DB for ; Mon, 25 Jan 2021 19:19:17 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id B935220E65 for ; Mon, 25 Jan 2021 19:19:16 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org B935220E65 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74339.133616 (Exim 4.92) (envelope-from ) id 1l47O5-0001qd-1M; Mon, 25 Jan 2021 19:18:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74339.133616; Mon, 25 Jan 2021 19:18:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47O4-0001qQ-Ql; Mon, 25 Jan 2021 19:18:56 +0000 Received: by outflank-mailman (input) for mailman id 74339; Mon, 25 Jan 2021 19:18:55 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47GE-0008N7-JC for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:50 +0000 Received: from mail-wr1-x42a.google.com (unknown [2a00:1450:4864:20::42a]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 4795a1b6-9277-4251-a093-50267805e128; Mon, 25 Jan 2021 19:09:19 +0000 (UTC) Received: by mail-wr1-x42a.google.com with SMTP id a9so14104888wrt.5 for ; Mon, 25 Jan 2021 11:09:16 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.14 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:14 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 4795a1b6-9277-4251-a093-50267805e128 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=YVKNotkZP3LGsaVlxkSD6nepQirnWE4Ml/8fGs0AunA=; b=KrvcTpK5//lShtuTFbhPnUDrnOYuCmGUPsDYI08ZNlDAcEmg9rEkmK1rqsje0pLvPG YW7ecdpI8aY9r+gHLknA9TQF5G0ValzbZIDhpG+BxcIp6DSXW4lR5jnLlwCE2k3mbOEF QKLP6idr15bErxFkIPbzIu8n1/xoGKVKx4C8/JFCS3LnDFPOneW23djVW2gSUW5JRnJi 83S2ENZrzQb5Ix7jl7uFFSiES67NGvcIx7nn0YpWzytDOblfawc7Y1V05tj+lZo88jLK 4eDKp+Tm9boGN1ViypRI5EH1zgslBVfd7CLb2OrAlJn4hi+k1/4NT6o/2Zz/A/JYLq6H LLkA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=YVKNotkZP3LGsaVlxkSD6nepQirnWE4Ml/8fGs0AunA=; b=WQMVyqSUWZBy4uJNtz09j0sOOO5XcKIlqjYaVoH/3FbJl/oijPjlN1cdUB+s/nWf0z iltsdZlxDDWxHmk/0sxWqeX6mt3j/UllnW0FOlSlZxeeogYjypeZCCI/DhAoOaxmbrSn yl7q1ppgZcdXc+397M/dqnckbPPQaEgcWD+baYZTHzxxgDwDZsyXgV5a7LUgo0Dnstii Z4rcnlibcf7dvSyPZvzTeGRnc/CEQkQWzeIo3Xs2ejgf5iy1UkkBD24NNgxxo2oeNMxh LUEZkTN5eEzcqez+N03Qdd90kg3aKnAVUq6Hlvl3OgHpub9UHDVtm7qDSPcIYp2uv1Ay WABg== X-Gm-Message-State: AOAM533kf81+20HefHm2iYbvK+xzC9OmNhIeuFyMpbqBHOIMeGd62Qra NGGBj51KLqos9fz0ZVHefGrPA+eRyyk5EQ== X-Google-Smtp-Source: ABdhPJx6ul0IDUGQLSGjUmZGGbbFGy2mcCuPZmlA6slGbF+nvYZjj2uDsI32Pm03xQR6Z1gfsqB5aQ== X-Received: by 2002:a5d:4806:: with SMTP id l6mr2558481wrq.389.1611601754995; Mon, 25 Jan 2021 11:09:14 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Jan Beulich , Andrew Cooper , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Wei Liu , George Dunlap , Ian Jackson , Julien Grall , Stefano Stabellini , Paul Durrant , Julien Grall Subject: [PATCH V5 21/22] xen/ioreq: Make x86's send_invalidate_req() common Date: Mon, 25 Jan 2021 21:08:28 +0200 Message-Id: <1611601709-28361-22-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko As the IOREQ is a common feature now and we also need to invalidate qemu/demu mapcache on Arm when the required condition occurs this patch moves this function to the common code (and remames it to ioreq_signal_mapcache_invalidate). This patch also moves per-domain qemu_mapcache_invalidate variable out of the arch sub-struct (and drops "qemu" prefix). We don't put this variable inside the #ifdef CONFIG_IOREQ_SERVER at the end of struct domain, but in the hole next to the group of 5 bools further up which is more efficient. The subsequent patch will add mapcache invalidation handling on Arm. Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Reviewed-by: Paul Durrant Acked-by: Jan Beulich [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" Changes RFC -> V1: - move send_invalidate_req() to the common code - update patch subject/description - move qemu_mapcache_invalidate out of the arch sub-struct, update checks - remove #if defined(CONFIG_ARM64) from the common code Changes V1 -> V2: - was split into: - xen/ioreq: Make x86's send_invalidate_req() common - xen/arm: Add mapcache invalidation handling - update patch description/subject - move Arm bits to a separate patch - don't alter the common code, the flag is set by arch code - rename send_invalidate_req() to send_invalidate_ioreq() - guard qemu_mapcache_invalidate with CONFIG_IOREQ_SERVER - use bool instead of bool_t - remove blank line blank line between head comment and #include-s Changes V2 -> V3: - update patch description - drop "qemu" prefix from the variable name - rename send_invalidate_req() to ioreq_signal_mapcache_invalidate() Changes V3 -> V4: - change variable location in struct domain Changes V4 -> V5: - add Jan's A-b and Paul's R-b --- --- xen/arch/x86/hvm/hypercall.c | 9 +++++---- xen/arch/x86/hvm/io.c | 14 -------------- xen/common/ioreq.c | 14 ++++++++++++++ xen/include/asm-x86/hvm/domain.h | 1 - xen/include/asm-x86/hvm/io.h | 1 - xen/include/xen/ioreq.h | 1 + xen/include/xen/sched.h | 5 +++++ 7 files changed, 25 insertions(+), 20 deletions(-) diff --git a/xen/arch/x86/hvm/hypercall.c b/xen/arch/x86/hvm/hypercall.c index ac573c8..6d41c56 100644 --- a/xen/arch/x86/hvm/hypercall.c +++ b/xen/arch/x86/hvm/hypercall.c @@ -20,6 +20,7 @@ */ #include #include +#include #include #include @@ -47,7 +48,7 @@ static long hvm_memory_op(int cmd, XEN_GUEST_HANDLE_PARAM(void) arg) rc = compat_memory_op(cmd, arg); if ( (cmd & MEMOP_CMD_MASK) == XENMEM_decrease_reservation ) - curr->domain->arch.hvm.qemu_mapcache_invalidate = true; + curr->domain->mapcache_invalidate = true; return rc; } @@ -326,9 +327,9 @@ int hvm_hypercall(struct cpu_user_regs *regs) HVM_DBG_LOG(DBG_LEVEL_HCALL, "hcall%lu -> %lx", eax, regs->rax); - if ( unlikely(currd->arch.hvm.qemu_mapcache_invalidate) && - test_and_clear_bool(currd->arch.hvm.qemu_mapcache_invalidate) ) - send_invalidate_req(); + if ( unlikely(currd->mapcache_invalidate) && + test_and_clear_bool(currd->mapcache_invalidate) ) + ioreq_signal_mapcache_invalidate(); return curr->hcall_preempted ? HVM_HCALL_preempted : HVM_HCALL_completed; } diff --git a/xen/arch/x86/hvm/io.c b/xen/arch/x86/hvm/io.c index 66a37ee..046a8eb 100644 --- a/xen/arch/x86/hvm/io.c +++ b/xen/arch/x86/hvm/io.c @@ -64,20 +64,6 @@ void send_timeoffset_req(unsigned long timeoff) gprintk(XENLOG_ERR, "Unsuccessful timeoffset update\n"); } -/* Ask ioemu mapcache to invalidate mappings. */ -void send_invalidate_req(void) -{ - ioreq_t p = { - .type = IOREQ_TYPE_INVALIDATE, - .size = 4, - .dir = IOREQ_WRITE, - .data = ~0UL, /* flush all */ - }; - - if ( ioreq_broadcast(&p, false) != 0 ) - gprintk(XENLOG_ERR, "Unsuccessful map-cache invalidate\n"); -} - bool hvm_emulate_one_insn(hvm_emulate_validate_t *validate, const char *descr) { struct hvm_emulate_ctxt ctxt; diff --git a/xen/common/ioreq.c b/xen/common/ioreq.c index 5b0f03e..67ef1f7 100644 --- a/xen/common/ioreq.c +++ b/xen/common/ioreq.c @@ -35,6 +35,20 @@ #include #include +/* Ask ioemu mapcache to invalidate mappings. */ +void ioreq_signal_mapcache_invalidate(void) +{ + ioreq_t p = { + .type = IOREQ_TYPE_INVALIDATE, + .size = 4, + .dir = IOREQ_WRITE, + .data = ~0UL, /* flush all */ + }; + + if ( ioreq_broadcast(&p, false) != 0 ) + gprintk(XENLOG_ERR, "Unsuccessful map-cache invalidate\n"); +} + static void set_ioreq_server(struct domain *d, unsigned int id, struct ioreq_server *s) { diff --git a/xen/include/asm-x86/hvm/domain.h b/xen/include/asm-x86/hvm/domain.h index b8be1ad..cf959f6 100644 --- a/xen/include/asm-x86/hvm/domain.h +++ b/xen/include/asm-x86/hvm/domain.h @@ -122,7 +122,6 @@ struct hvm_domain { struct viridian_domain *viridian; - bool_t qemu_mapcache_invalidate; bool_t is_s3_suspended; /* diff --git a/xen/include/asm-x86/hvm/io.h b/xen/include/asm-x86/hvm/io.h index 6bc80db..5eb4ddb 100644 --- a/xen/include/asm-x86/hvm/io.h +++ b/xen/include/asm-x86/hvm/io.h @@ -97,7 +97,6 @@ bool relocate_portio_handler( unsigned int size); void send_timeoffset_req(unsigned long timeoff); -void send_invalidate_req(void); bool handle_mmio_with_translation(unsigned long gla, unsigned long gpfn, struct npfec); bool handle_pio(uint16_t port, unsigned int size, int dir); diff --git a/xen/include/xen/ioreq.h b/xen/include/xen/ioreq.h index 89ee171..2d635e9 100644 --- a/xen/include/xen/ioreq.h +++ b/xen/include/xen/ioreq.h @@ -103,6 +103,7 @@ struct ioreq_server *ioreq_server_select(struct domain *d, int ioreq_send(struct ioreq_server *s, ioreq_t *proto_p, bool buffered); unsigned int ioreq_broadcast(ioreq_t *p, bool buffered); +void ioreq_signal_mapcache_invalidate(void); void ioreq_domain_init(struct domain *d); diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index 59e5b6a..06dba1a 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -444,6 +444,11 @@ struct domain * unpaused for the first time by the systemcontroller. */ bool creation_finished; + /* + * Indicates that mapcache invalidation request should be sent to + * the device emulator. + */ + bool mapcache_invalidate; /* Which guest this guest has privileges on */ struct domain *target; From patchwork Mon Jan 25 19:08:29 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oleksandr Tyshchenko X-Patchwork-Id: 12044097 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1E78DC433DB for ; Mon, 25 Jan 2021 19:21:18 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id B5FAA2256F for ; Mon, 25 Jan 2021 19:21:17 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org B5FAA2256F Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from list by lists.xenproject.org with outflank-mailman.74381.133745 (Exim 4.92) (envelope-from ) id 1l47QD-0003fu-0X; Mon, 25 Jan 2021 19:21:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 74381.133745; Mon, 25 Jan 2021 19:21:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47QC-0003fU-LX; Mon, 25 Jan 2021 19:21:08 +0000 Received: by outflank-mailman (input) for mailman id 74381; Mon, 25 Jan 2021 19:21:07 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1l47GJ-0008N7-JR for xen-devel@lists.xenproject.org; Mon, 25 Jan 2021 19:10:55 +0000 Received: from mail-wm1-x333.google.com (unknown [2a00:1450:4864:20::333]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id 5bc4c605-df07-442b-8357-7150c8be4ed9; Mon, 25 Jan 2021 19:09:19 +0000 (UTC) Received: by mail-wm1-x333.google.com with SMTP id m187so388585wme.2 for ; Mon, 25 Jan 2021 11:09:16 -0800 (PST) Received: from otyshchenko.www.tendawifi.com ([212.22.223.21]) by smtp.gmail.com with ESMTPSA id k6sm12991031wro.27.2021.01.25.11.09.15 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 25 Jan 2021 11:09:15 -0800 (PST) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: 5bc4c605-df07-442b-8357-7150c8be4ed9 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=56q3vS7kTPjkBeZpCedlA35sj1F6SAbFf73YA28cuDw=; b=NAQXLR3YC0ePDFcJsMH2mJiS9iq1e7hqncXyhKPgAIZMXHKYL90qKiw9NPln14JyhJ ko/xbh3dydPY5irT9V1He2YZE32F2nM/EXck5OgfqpTzh9mC+T43vXlsCduaHjZM+u8/ CMyvM8Vsx81GVsTyVd1aQeaYP9LO7AekDaITxS+WU/0KejucOPTaVBZlwGLMw2dG8oQY ACi9SUyne7kX/xeLPKxVZ6my/XbN0GiGE1c9eA6iYAOBRSd+gXd/cnmqexzyejRGTeYG Imara8PFomv2OPOgIUkvSHWHkOUSc4zIxe926xQqwcDHpS2RyOHOeTEkSE6D/6ncvsd2 00AQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=56q3vS7kTPjkBeZpCedlA35sj1F6SAbFf73YA28cuDw=; b=teMn0cPjRhsuaOc0sTZQXvngmm8sK4YvDwvCt00FViUcC6D8WVxYDGhoGOH77WrekW CFYGCIP+IiBOPBpssnT7eBOlYGOWVhyjdT+BIn9NPEaDiqZgwNuelKzS1ezaSAbQq951 8f7z0V4IffMGznwXgg1cGYlCkZCfUVm0IY6QbP3eOTBQdHPpqKfot93xP3r/RjYpU7ZD DAZe/ISlJGG1desqYtf2OLHEQfBi5935JwrjOcX8XFS/6dORFGZ2rnwE6xR716kFdDXE LSRWC2FgZx4R8DjhaUhwf7MQx/W/OqvKXhp83rmgkhQw2CUaB88J4Yru7xpkQMmBDJx9 /Jqw== X-Gm-Message-State: AOAM530l4gmylGvep/xpp00r7krgexN+w1CNpJFkGPXCStpuFZuInwoh Wj9QKVTLAkeSBCusKq6f1p4WZxwQjninRA== X-Google-Smtp-Source: ABdhPJxqWxZA1UNJHCs8azX6BOxbaVbtxb6CX9NTt1ZK8DMoPsOuUD+DEbXrlAAjS6M6ec1aRQWp0Q== X-Received: by 2002:a1c:a593:: with SMTP id o141mr1467438wme.92.1611601755965; Mon, 25 Jan 2021 11:09:15 -0800 (PST) From: Oleksandr Tyshchenko To: xen-devel@lists.xenproject.org Cc: Oleksandr Tyshchenko , Stefano Stabellini , Julien Grall , Volodymyr Babchuk , Julien Grall Subject: [PATCH V5 22/22] xen/arm: Add mapcache invalidation handling Date: Mon, 25 Jan 2021 21:08:29 +0200 Message-Id: <1611601709-28361-23-git-send-email-olekstysh@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> References: <1611601709-28361-1-git-send-email-olekstysh@gmail.com> From: Oleksandr Tyshchenko We need to send mapcache invalidation request to qemu/demu everytime the page gets removed from a guest. At the moment, the Arm code doesn't explicitely remove the existing mapping before inserting the new mapping. Instead, this is done implicitely by __p2m_set_entry(). So we need to recognize a case when old entry is a RAM page *and* the new MFN is different in order to set the corresponding flag. The most suitable place to do this is p2m_free_entry(), there we can find the correct leaf type. The invalidation request will be sent in do_trap_hypercall() later on. Taking into the account the following the do_trap_hypercall() is the best place to send invalidation request: - The only way a guest can modify its P2M on Arm is via an hypercall - When sending the invalidation request, the vCPU will be blocked until all the IOREQ servers have acknowledged the invalidation Signed-off-by: Oleksandr Tyshchenko CC: Julien Grall Reviewed-by: Stefano Stabellini [On Arm only] Tested-by: Wei Chen --- Please note, this is a split/cleanup/hardening of Julien's PoC: "Add support for Guest IO forwarding to a device emulator" *** Please note, this patch depends on the following which is on review: https://patchwork.kernel.org/patch/11803383/ This patch is on par with x86 code (whether it is buggy or not). If there is a need to improve/harden something, this can be done on a follow-up. *** Changes V1 -> V2: - new patch, some changes were derived from (+ new explanation): xen/ioreq: Make x86's invalidate qemu mapcache handling common - put setting of the flag into __p2m_set_entry() - clarify the conditions when the flag should be set - use domain_has_ioreq_server() - update do_trap_hypercall() by adding local variable Changes V2 -> V3: - update patch description - move check to p2m_free_entry() - add a comment - use "curr" instead of "v" in do_trap_hypercall() Changes V3 -> V4: - update patch description - re-order check in p2m_free_entry() to call domain_has_ioreq_server() only if p2m->domain == current->domain - add a comment in do_trap_hypercall() Changes V4 -> V5: - add Stefano's R-b - update comment in do_trap_hypercall() --- --- xen/arch/arm/p2m.c | 25 +++++++++++++++++-------- xen/arch/arm/traps.c | 20 +++++++++++++++++--- 2 files changed, 34 insertions(+), 11 deletions(-) diff --git a/xen/arch/arm/p2m.c b/xen/arch/arm/p2m.c index d41c4fa..26acb95d 100644 --- a/xen/arch/arm/p2m.c +++ b/xen/arch/arm/p2m.c @@ -1,6 +1,7 @@ #include #include #include +#include #include #include #include @@ -749,17 +750,25 @@ static void p2m_free_entry(struct p2m_domain *p2m, if ( !p2m_is_valid(entry) ) return; - /* Nothing to do but updating the stats if the entry is a super-page. */ - if ( p2m_is_superpage(entry, level) ) + if ( p2m_is_superpage(entry, level) || (level == 3) ) { - p2m->stats.mappings[level]--; - return; - } +#ifdef CONFIG_IOREQ_SERVER + /* + * If this gets called (non-recursively) then either the entry + * was replaced by an entry with a different base (valid case) or + * the shattering of a superpage was failed (error case). + * So, at worst, the spurious mapcache invalidation might be sent. + */ + if ( (p2m->domain == current->domain) && + domain_has_ioreq_server(p2m->domain) && + p2m_is_ram(entry.p2m.type) ) + p2m->domain->mapcache_invalidate = true; +#endif - if ( level == 3 ) - { p2m->stats.mappings[level]--; - p2m_put_l3_page(entry); + /* Nothing to do if the entry is a super-page. */ + if ( level == 3 ) + p2m_put_l3_page(entry); return; } diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c index 4cdd343..64b740b 100644 --- a/xen/arch/arm/traps.c +++ b/xen/arch/arm/traps.c @@ -1443,6 +1443,7 @@ static void do_trap_hypercall(struct cpu_user_regs *regs, register_t *nr, const union hsr hsr) { arm_hypercall_fn_t call = NULL; + struct vcpu *curr = current; BUILD_BUG_ON(NR_hypercalls < ARRAY_SIZE(arm_hypercall_table) ); @@ -1459,7 +1460,7 @@ static void do_trap_hypercall(struct cpu_user_regs *regs, register_t *nr, return; } - current->hcall_preempted = false; + curr->hcall_preempted = false; perfc_incra(hypercalls, *nr); call = arm_hypercall_table[*nr].fn; @@ -1472,7 +1473,7 @@ static void do_trap_hypercall(struct cpu_user_regs *regs, register_t *nr, HYPERCALL_RESULT_REG(regs) = call(HYPERCALL_ARGS(regs)); #ifndef NDEBUG - if ( !current->hcall_preempted ) + if ( !curr->hcall_preempted ) { /* Deliberately corrupt parameter regs used by this hypercall. */ switch ( arm_hypercall_table[*nr].nr_args ) { @@ -1489,8 +1490,21 @@ static void do_trap_hypercall(struct cpu_user_regs *regs, register_t *nr, #endif /* Ensure the hypercall trap instruction is re-executed. */ - if ( current->hcall_preempted ) + if ( curr->hcall_preempted ) regs->pc -= 4; /* re-execute 'hvc #XEN_HYPERCALL_TAG' */ + +#ifdef CONFIG_IOREQ_SERVER + /* + * We call ioreq_signal_mapcache_invalidate from do_trap_hypercall() + * because the only way a guest can modify its P2M on Arm is via an + * hypercall. + * Note that sending the invalidation request causes the vCPU to block + * until all the IOREQ servers have acknowledged the invalidation. + */ + if ( unlikely(curr->domain->mapcache_invalidate) && + test_and_clear_bool(curr->domain->mapcache_invalidate) ) + ioreq_signal_mapcache_invalidate(); +#endif } void arch_hypercall_tasklet_result(struct vcpu *v, long res)